fix: wip mem xf

.gitignore

@@ -1,3 +1,8 @@
+*.arrow
+*.swp
+# ignore knn index
+gpt4all/index/**/
+.DS_Store
 *.pkl
 ckpts*
 .deepspeed_env

configs/deepspeed/ds_config_gptj.json

@@ -35,14 +35,5 @@
 		  ],
 		  "eps": 1e-08
 		}
-	  },
-	  "scheduler": {
-		"type": "WarmupLR",
-		"params": {
-		  "warmup_min_lr": 0,
-		  "warmup_max_lr": "auto",
-		  "warmup_num_steps": "auto",
-		  "warmup_type": "linear"
-		}
 	  }
 }

configs/deepspeed/ds_config_pythiaseek.json

@@ -0,0 +1,39 @@
+{
+	"train_batch_size": "auto",
+	"gradient_accumulation_steps": "auto",
+	"train_micro_batch_size_per_gpu": "auto",
+	"fp16": {
+	  "enabled": "auto",
+	  "min_loss_scale": 1,
+	  "loss_scale_window": 1000,
+	  "hysteresis": 2,
+	  "initial_scale_power": 32
+	},
+	"bf16": {
+		"enabled": "auto"
+	},
+	"gradient_clipping": 1.0,
+	"zero_optimization": {
+	  "stage": 2,
+	  "offload_param": {
+		"device": "none"
+	  },
+	  "offload_optimizer": {
+		"device": "none"
+	  },
+	  "allgather_partitions": true,
+	  "allgather_bucket_size": 5e8,
+	  "contiguous_gradients": true
+	},
+	"optimizer": {
+		"type": "AdamW",
+		"params": {
+		  "lr": "auto",
+		  "betas": [
+			0.9,
+			0.999
+		  ],
+		  "eps": 1e-08
+		}
+	  }
+}

configs/eval/evaluate_gpt4all_jr.yaml

@@ -0,0 +1,18 @@
+# model/tokenizer
+model_name: "/home/paperspace/gpt4all/gpt4all/train/ckpts/epoch_2"
+tokenizer_name: "EleutherAI/gpt-j-6B"
+version: null
+gradient_checkpointing: true
+save_name: "nomic-ai/gpt-jr"
+encoder_dim: 384
+
+# dataset
+streaming: false
+num_proc: 64
+dataset_path: "/home/paperspace/gpt4all/gpt4all/index/squad_supplemented_validation" 
+max_length: 1024
+batch_size: 32
+pct_test: 0.05
+q_column: "question"
+a_column: "answers"
+encoder_column: "neighbor_embeddings"

configs/eval/evaluate_lethe.yaml

@@ -0,0 +1,23 @@
+# model/tokenizer
+model_name: "/home/paperspace/gpt4all/gpt4all/train/ckpts/mem_attn/step_1000"
+tokenizer_name: "EleutherAI/pythia-1b"
+version: null
+gradient_checkpointing: false
+memory_attn_layer: 12
+
+
+# dataset
+streaming: false
+num_proc: 64
+dataset_path: "/home/paperspace/gpt4all/gpt4all/inference/synth_data_combined_174"
+# dataset_path: "/home/paperspace/gpt4all/gpt4all/index/squad_supplemented_validation" 
+max_length: 1024
+batch_size: 1
+pct_test: 0.05
+q_column: "question"
+a_column: "answer"
+context_column: "text"
+num_memories_per_index: 2000000
+num_neighbors_to_retrieve: 2
+num_neighbors_to_store: 1 
+mem_chunk_size: 64

configs/eval/evaluate_pythiaseek.yaml

@@ -0,0 +1,18 @@
+# model/tokenizer
+model_name: "/home/paperspace/gpt4all/gpt4all/train/ckpts/learnable_alpha/epoch_0"
+tokenizer_name: "EleutherAI/pythia-1b"
+version: null
+gradient_checkpointing: true
+save_name: "nomic-ai/gpt-jr"
+encoder_dim: 384
+
+# dataset
+streaming: false
+num_proc: 64
+dataset_path: "/home/paperspace/gpt4all/gpt4all/index/squad_supplemented_validation" 
+max_length: 1024
+batch_size: 32
+pct_test: 0.05
+q_column: "question"
+a_column: "answers"
+encoder_column: "neighbor_embeddings"

configs/index/finetune_gptjr.yaml

@@ -0,0 +1,13 @@
+# dataset
+streaming: false
+num_proc: 64
+dataset_path: "squad"
+max_length: 1024
+batch_size: 32
+
+#index
+index_path: "/home/paperspace/gpt4all/gpt4all/index/wiki-cohere-index.bin"
+index_database: "nomic-ai/cohere-wiki-sbert"
+index_space: "cosine"
+index_dim: 384
+query_embedding_field: 'question'

configs/inference/synth_data.yaml

@@ -0,0 +1,19 @@
+model_name: "nomic-ai/gpt4all-j"
+revision: "v1.3-groovy"
+tokenizer_name: "nomic-ai/gpt4all-j"
+
+# dataset
+streaming: false
+num_proc: 64
+dataset_path: "nomic-ai/cohere-wiki-sbert"
+batch_size: 32 
+output_path: "synth_qa_pairs"
+
+# generation
+max_new_tokens: 75
+max_generations: 200000
+
+save_every: 1000
+
+
+seed: 42

configs/train/finetune_gptjr.yaml

@@ -0,0 +1,43 @@
+# model/tokenizer
+model_name: "EleutherAI/gpt-j-6b"
+tokenizer_name: "EleutherAI/gpt-j-6b"
+version: null
+gradient_checkpointing: true
+save_name: "nomic-ai/gpt-jr-decay-alpha"
+push_to_hub: false
+encoder_dim: 384
+
+# dataset
+streaming: false
+num_proc: 64
+dataset_path: "/home/paperspace/gpt4all/gpt4all/index/squad_supplemented_train" 
+max_length: 1024
+batch_size: 8
+pct_test: 0.05
+q_column: "question"
+a_column: "answers"
+encoder_column: "neighbor_embeddings"
+
+
+# train dynamics
+lr: 1.0e-4
+min_lr: 0 
+weight_decay: 0.0
+eval_every: 50 
+save_every: -1
+log_grads_every: 100
+log_lr_every: 10
+output_dir: "ckpts/decay_alpha"
+checkpoint: null
+lora: false
+warmup_steps: 500
+num_epochs: 5 
+debug: false
+scheduler: false
+
+# logging
+wandb: true
+wandb_entity: gpt4all
+wandb_project_name: retrieval
+seed: 42
+

configs/train/finetune_memory.yaml

@@ -0,0 +1,46 @@
+# model/tokenizer
+model_name: "EleutherAI/pythia-1b"
+tokenizer_name: "EleutherAI/pythia-1b"
+version: null
+gradient_checkpointing: true
+save_name: "nomic-ai/lethe"
+push_to_hub: false
+memory_attn_layer: 12
+
+# dataset
+streaming: false
+num_proc: 64
+dataset_path: "/home/paperspace/gpt4all/gpt4all/inference/synth_data_combined_174"
+max_length: 1024
+batch_size: 32
+pct_test: 0.05
+q_column: "question"
+a_column: "answer"
+context_column: "text"
+num_memories_per_index: 2000000
+num_neighbors_to_retrieve: 2
+num_neighbors_to_store: 1 
+mem_chunk_size: 64
+
+# train dynamics
+lr: 1.0e-5
+min_lr: 0 
+weight_decay: 0.0
+eval_every: 100 
+save_every: 100
+log_grads_every: 100
+log_lr_every: 10
+output_dir: "ckpts/mem_attn_no_cosine_sim"
+checkpoint: null
+lora: false
+warmup_steps: 200
+num_epochs: 5 
+debug: false
+scheduler: false
+
+# logging
+wandb: true
+wandb_entity: gpt4all
+wandb_project_name: mem_attn
+seed: 42
+

configs/train/finetune_pythiaseek.yaml

@@ -0,0 +1,46 @@
+# model/tokenizer
+model_name: "EleutherAI/pythia-1b"
+tokenizer_name: "EleutherAI/pythia-1b"
+version: null
+gradient_checkpointing: true
+save_name: "nomic-ai/pythiaseek-large-bs"
+push_to_hub: false
+encoder_dim: 384
+learnable_alpha: true
+cross_attn_layer: 12
+freeze_pretrained: false
+
+# dataset
+streaming: false
+num_proc: 64
+dataset_path: "/home/paperspace/gpt4all/gpt4all/index/squad_supplemented_train" 
+max_length: 1024
+batch_size: 64
+pct_test: 0.05
+q_column: "question"
+a_column: "answers"
+encoder_column: "neighbor_embeddings"
+
+
+# train dynamics
+lr: 1.0e-4
+min_lr: 0 
+weight_decay: 0.0
+eval_every: 100 
+save_every: -1
+log_grads_every: 100
+log_lr_every: 10
+output_dir: "ckpts/learnable_alpha_6b"
+checkpoint: null
+lora: false
+warmup_steps: 500
+num_epochs: 5 
+debug: false
+scheduler: false
+
+# logging
+wandb: true
+wandb_entity: gpt4all
+wandb_project_name: retrieval
+seed: 42
+

configs/train/pretrain_minipile.yaml

@@ -0,0 +1,43 @@
+# model/tokenizer
+model_name: "EleutherAI/pythia-1b"
+tokenizer_name: "EleutherAI/pythia-1b"
+version: null
+gradient_checkpointing: true
+save_name: "nomic-ai/minipille"
+push_to_hub: false
+memory_attn_layer: 12
+
+# dataset
+streaming: false
+num_proc: 64
+dataset_path: "JeanKaddour/minipile" 
+max_length: 2048
+batch_size: 64
+pct_test: 0.05
+num_memories_per_index: 5000000
+mem_chunk_size: 512
+num_chunks: 10
+num_neighbors_to_retrieve: 32
+
+# train dynamics
+lr: 1.0e-4
+min_lr: 0 
+weight_decay: 0.0
+eval_every: 100 
+save_every: -1
+log_grads_every: 100
+log_lr_every: 10
+output_dir: "ckpts/minipile"
+checkpoint: null
+lora: false
+warmup_steps: 500
+num_epochs: 5 
+debug: false
+scheduler: false
+
+# logging
+wandb: true
+wandb_entity: gpt4all
+wandb_project_name: minipile
+seed: 42
+

create_hostname.sh

@@ -1,8 +0,0 @@
-#!/bin/bash
-
-export WORKER_IP=$1
-N_GPUS=8
-# create dir if doesn't exist
-sudo mkdir -p /job
-printf "localhost slots=$N_GPUS\n$WORKER_IP slots=$N_GPUS" | sudo tee /job/hostfile
-echo /job/hostfile

gpt4all/data/__init__.py

@@ -0,0 +1,2 @@
+from .instruction_tuning_dataloader import *
+from .retrieval_dataloader import *

gpt4all/data/instruction_tuning_dataloader.py

@@ -1,61 +1,11 @@
 import glob
 import torch
-from datasets import load_dataset, concatenate_datasets
+from datasets import load_dataset
 import os
+import hnswlib
 from torch.utils.data import DataLoader
 from transformers import DefaultDataCollator
-
-
-
-def tokenize_inputs(config, tokenizer, examples):
-    max_length = config["max_length"]
-
-    # hacky backward compatible
-    different_eos = tokenizer.eos_token != "</s>"
-    out = {"labels": [], "input_ids": []}
-    for prompt, response in zip(examples["prompt"], examples["response"]):
-        if different_eos:
-            if response.count("</s> \n") > 0:
-                response = response.replace("</s> \n", f"{tokenizer.eos_token} \n") 
-
-        prompt_len = len(tokenizer(prompt + "\n", return_tensors="pt")["input_ids"][0])
-
-        # hack if our prompt is super long
-        # we need to include some labels so we arbitrarily trunacate at max_length // 2
-        # if the length is too long
-        if prompt_len >= max_length // 2:
-            # if prompt is too long, truncate
-            # but make sure to truncate to at max 1024 tokens
-            new_len = min(max_length // 2, len(prompt) // 2)
-            prompt = prompt[:new_len]
-            # get new prompt length
-            prompt_len = tokenizer(prompt + "\n", return_tensors="pt", max_length=max_length // 2, truncation=True).input_ids.ne(tokenizer.pad_token_id).sum().item()
-
-        assert prompt_len <= max_length // 2, f"prompt length {prompt_len} exceeds max length {max_length}"
-
-        input_tokens = tokenizer(prompt + "\n" + response + tokenizer.eos_token,
-                                 truncation=True, max_length=max_length, return_tensors="pt")["input_ids"].squeeze()
-
-        labels = input_tokens.clone()
-        labels[:prompt_len] = -100
-        if len(labels) < max_length:
-            # pad to max_length with -100
-            labels = torch.cat([labels, torch.full((max_length - len(labels),), -100)])
-
-        assert (labels == -100).sum() < len(labels), f"Labels are all -100, something wrong. prompt length {prompt_len} exceeds max length {max_length}" 
-        
-        if (labels == -100).sum() == len(labels) - 1:
-            print(prompt)
-            print(response)
-            raise
-
-        input_tokens = tokenizer.pad({"input_ids": input_tokens}, padding="max_length", max_length=max_length)["input_ids"]
-        out["labels"].append(labels)
-        out["input_ids"].append(input_tokens)
-
-    out = {k: torch.stack(v) if isinstance(v, list) else v for k, v in out.items()}
-
-    return out
+from .preprocess import tokenize_inputs
 
 
 def load_data(config, tokenizer):
@@ -85,13 +35,13 @@ def load_data(config, tokenizer):
 
     # tokenize inputs and return labels and attention mask
     train_dataset = train_dataset.map(
-        lambda ele: tokenize_inputs(config, tokenizer, ele),
+        lambda ele: tokenize_inputs(config, tokenizer, ele, "prompt", "response"),
         batched=True,
         remove_columns=["source", "prompt"],
         **kwargs
     )
     val_dataset = val_dataset.map(
-        lambda ele: tokenize_inputs(config, tokenizer, ele),
+        lambda ele: tokenize_inputs(config, tokenizer, ele, "prompt", "response"),
         batched=True,
         remove_columns=["source", "prompt"],
         **kwargs
@@ -116,6 +66,7 @@ def load_data(config, tokenizer):
 
     return train_dataloader, val_dataloader
 
+
     
 def load_data_for_inference(config, tokenizer):
     dataset_path = config["dataset_path"]
@@ -152,12 +103,12 @@ def load_data_for_inference(config, tokenizer):
 
     # tokenize inputs and return labels and attention mask
     train_dataset = train_dataset.map(
-        lambda ele: tokenize_inputs(config, tokenizer, ele),
+        lambda ele: tokenize_inputs(config, tokenizer, ele, "prompt", "response"),
         batched=True,
         **kwargs
     )
     val_dataset = val_dataset.map(
-        lambda ele: tokenize_inputs(config, tokenizer, ele), 
+        lambda ele: tokenize_inputs(config, tokenizer, ele, "prompt", "response"), 
         batched=True,
         **kwargs
     )

gpt4all/data/preprocess.py

@@ -0,0 +1,51 @@
+import torch
+
+def tokenize_inputs(config, tokenizer, examples, input_col, target_col):
+    max_length = config["max_length"]
+
+    # hacky backward compatible
+    different_eos = tokenizer.eos_token != "</s>"
+    out = {"labels": [], "input_ids": []}
+    for prompt, response in zip(examples[input_col], examples[target_col]):
+        if different_eos:
+            if response.count("</s> \n") > 0:
+                response = response.replace("</s> \n", f"{tokenizer.eos_token} \n") 
+
+        prompt_len = len(tokenizer(prompt + "\n", return_tensors="pt")["input_ids"][0])
+
+        # hack if our prompt is super long
+        # we need to include some labels so we arbitrarily trunacate at max_length // 2
+        # if the length is too long
+        if prompt_len >= max_length // 2:
+            # if prompt is too long, truncate
+            # but make sure to truncate to at max 1024 tokens
+            new_len = min(max_length // 2, len(prompt) // 2)
+            prompt = prompt[:new_len]
+            # get new prompt length
+            prompt_len = tokenizer(prompt + "\n", return_tensors="pt", max_length=max_length // 2, truncation=True).input_ids.ne(tokenizer.pad_token_id).sum().item()
+
+        assert prompt_len <= max_length // 2, f"prompt length {prompt_len} exceeds max length {max_length}"
+
+        input_tokens = tokenizer(prompt + "\n" + response + tokenizer.eos_token,
+                                 truncation=True, max_length=max_length, return_tensors="pt")["input_ids"].squeeze()
+
+        labels = input_tokens.clone()
+        labels[:prompt_len] = -100
+        if len(labels) < max_length:
+            # pad to max_length with -100
+            labels = torch.cat([labels, torch.full((max_length - len(labels),), -100)])
+
+        assert (labels == -100).sum() < len(labels), f"Labels are all -100, something wrong. prompt length {prompt_len} exceeds max length {max_length}" 
+        
+        if (labels == -100).sum() == len(labels) - 1:
+            print(prompt)
+            print(response)
+            raise
+
+        input_tokens = tokenizer.pad({"input_ids": input_tokens}, padding="max_length", max_length=max_length)["input_ids"]
+        out["labels"].append(labels)
+        out["input_ids"].append(input_tokens)
+
+    out = {k: torch.stack(v) if isinstance(v, list) else v for k, v in out.items()}
+
+    return out

gpt4all/data/retrieval_dataloader.py

@@ -0,0 +1,203 @@
+from datasets import load_dataset, Dataset
+import os
+from torch.utils.data import DataLoader
+from .preprocess import tokenize_inputs
+from transformers import DefaultDataCollator
+
+
+def load_retrieval_augmented_data(config, tokenizer, split="train", split_dataset=True):
+    dataset_path = config["dataset_path"]
+
+    if os.path.exists(dataset_path):
+        dataset = Dataset.load_from_disk(dataset_path)
+    else:
+        dataset = load_dataset(dataset_path, split=split)
+
+
+    question_col = config["q_column"]
+    answer_col = config["a_column"]
+    encoder_column = config["encoder_column"]
+
+    if config["streaming"] is False:
+        kwargs = {"num_proc": config["num_proc"]}
+    else:
+        kwargs = {}
+
+    # strip any unneccessary whitespace
+    # there's one question that's includes a ton of whitespace
+    dataset = dataset.map(lambda ele: {question_col: [q.strip() for q in ele[question_col]]}, batched=True)
+    # in squad, the data is formatted where each ele in answers is a dict where the key text holds
+    # a list of the answer
+    dataset = dataset.map(lambda ele: {answer_col: [t["text"][0] for t in ele[answer_col]]}, batched=True)
+
+    dataset = dataset.map(
+        lambda ele: tokenize_inputs(config, tokenizer, ele, question_col, answer_col),
+        batched=True,
+        **kwargs
+    )
+
+    # tokenize inputs + labels in teacher-force format
+    # rename encoder hidden states if not already called that
+    if encoder_column != "encoder_hidden_states":
+        dataset = dataset.rename_column(encoder_column, "encoder_hidden_states")
+
+    columns_to_keep = ["input_ids", "labels", "encoder_hidden_states"]
+
+    col_names_to_rm = [col for col in dataset.column_names if col not in columns_to_keep]
+    dataset = dataset.remove_columns(col_names_to_rm)
+
+    if split_dataset:
+        dataset = dataset.train_test_split(test_size=config["pct_test"], seed=config["seed"])
+        train_dataset, val_dataset = dataset["train"], dataset["test"]
+
+        train_dataloader = DataLoader(
+            train_dataset,
+            batch_size=config["batch_size"],
+            collate_fn=DefaultDataCollator(),
+        )
+
+        val_dataloader = DataLoader(
+            val_dataset,
+            batch_size=config["batch_size"],
+            collate_fn=DefaultDataCollator(),
+        )
+
+        return train_dataloader, val_dataloader
+
+    else:
+        dataloader = DataLoader(
+            dataset,
+            batch_size=config["batch_size"],
+            collate_fn=DefaultDataCollator(),
+        )
+
+        return dataloader
+
+
+def load_memory_augmented_data(config, tokenizer, split="train", split_dataset=True):
+    dataset_path = config["dataset_path"]
+
+    if os.path.exists(dataset_path):
+        dataset = Dataset.load_from_disk(dataset_path)
+    else:
+        dataset = load_dataset(dataset_path, split=split)
+
+
+    question_col = config["q_column"]
+    answer_col = config["a_column"]
+    context_col = config["context_column"]
+
+    if config["streaming"] is False:
+        kwargs = {"num_proc": config["num_proc"]}
+    else:
+        kwargs = {}
+
+    # strip any unneccessary whitespace
+    # there's one question that's includes a ton of whitespace
+    dataset = dataset.map(lambda ele: {question_col: [q.strip() for q in ele[question_col]]}, batched=True)
+    # in squad, the data is formatted where each ele in answers is a dict where the key text holds
+    # a list of the answer
+    dataset = dataset.map(lambda ele: {answer_col: [t.strip() for t in ele[answer_col]]}, batched=True)
+    # dataset = dataset.map(lambda ele: {answer_col: [t["text"][0] for t in ele[answer_col]]}, batched=True)
+
+    dataset = dataset.map(
+        lambda ele: tokenize_inputs(config, tokenizer, ele, question_col, answer_col),
+        batched=True,
+        **kwargs
+    )
+
+    # tokenize contexts for each example
+    dataset = dataset.map(
+        lambda ele: {"retrieved_context": tokenizer([ele[context_col]], 
+                                                    return_tensors="pt", 
+                                                    padding="max_length", 
+                                                    truncation=True)["input_ids"]},
+        **kwargs
+    )
+
+    columns_to_keep = ["id", "input_ids", "labels", "retrieved_context"]
+
+    col_names_to_rm = [col for col in dataset.column_names if col not in columns_to_keep]
+    dataset = dataset.remove_columns(col_names_to_rm)
+
+    if split_dataset:
+        dataset = dataset.train_test_split(test_size=config["pct_test"], seed=config["seed"])
+        train_dataset, val_dataset = dataset["train"], dataset["test"]
+
+        train_dataloader = DataLoader(
+            train_dataset.remove_columns("id"),
+            batch_size=config["batch_size"],
+            collate_fn=DefaultDataCollator(),
+        )
+
+        val_dataloader = DataLoader(
+            val_dataset,
+            batch_size=config["batch_size"],
+            collate_fn=DefaultDataCollator(),
+        )
+
+        return train_dataloader, val_dataloader 
+
+    else:
+        dataloader = DataLoader(
+            dataset,
+            batch_size=config["batch_size"],
+            collate_fn=DefaultDataCollator(),
+        )
+
+        return dataloader
+
+
+def load_memory_pretraining_data(config, tokenizer, split="train", split_dataset=True):
+    dataset_path = config["dataset_path"]
+
+    if os.path.exists(dataset_path):
+        dataset = Dataset.load_from_disk(dataset_path)
+    else:
+        dataset = load_dataset(dataset_path, split=split)
+
+    if config["streaming"] is False:
+        kwargs = {"num_proc": config["num_proc"]}
+    else:
+        kwargs = {}
+        
+    # e.g. 512 * 10 = 5120 sequence length split up
+    max_length = config["mem_chunk_size"] * config["num_chunks"]
+    dataset = dataset.map(lambda ele: tokenizer(ele["text"], padding="max_length", truncation=True, max_length=max_length),
+                          batched=True, **kwargs)
+
+    dataset = dataset.map(lambda x: {"labels": x["input_ids"]}, batched=True, **kwargs)
+
+
+    columns_to_keep = ["input_ids", "labels", "attention_mask"]
+
+    col_names_to_rm = [col for col in dataset.column_names if col not in columns_to_keep]
+    dataset = dataset.remove_columns(col_names_to_rm)
+
+    # we can shuffle since the docs are in one row not split across rows
+    if split_dataset:
+        dataset = dataset.train_test_split(test_size=config["pct_test"], seed=config["seed"])
+        train_dataset, val_dataset = dataset["train"], dataset["test"]
+
+        train_dataloader = DataLoader(
+            train_dataset,
+            batch_size=config["batch_size"],
+            collate_fn=DefaultDataCollator(),
+        )
+
+        val_dataloader = DataLoader(
+            val_dataset,
+            batch_size=config["batch_size"],
+            collate_fn=DefaultDataCollator(),
+        )
+
+        return train_dataloader, val_dataloader
+
+    else:
+        dataloader = DataLoader(
+            dataset,
+            batch_size=config["batch_size"],
+            collate_fn=DefaultDataCollator(),
+        )
+
+        return dataloader

gpt4all/eval/eval_self_instruct.py

@@ -3,7 +3,7 @@ import torch
 import pickle
 import numpy as np
 from tqdm import tqdm
-from read import read_config
+from gpt4all.utils.read import read_config
 from argparse import ArgumentParser
 from peft import PeftModelForCausalLM
 from transformers import AutoModelForCausalLM, AutoTokenizer

gpt4all/eval/eval_squad.py

@@ -0,0 +1,59 @@
+import torch
+from gpt4all.models import GPTJRForCausalLM, PythiaSeekForCausalLM
+from gpt4all.data.retrieval_dataloader import load_retrieval_augmented_data
+from gpt4all.train.metrics import f1_score, exact_match_score
+from gpt4all.utils.read import read_config
+from transformers import AutoTokenizer
+from argparse import ArgumentParser
+from tqdm import tqdm
+
+parser = ArgumentParser()
+parser.add_argument("--config", type=str, default="config.yaml")
+
+args = parser.parse_args()
+
+config = read_config(args.config)
+
+tokenizer = AutoTokenizer.from_pretrained(config["tokenizer_name"])
+if tokenizer.pad_token is None:
+    tokenizer.pad_token = tokenizer.eos_token
+
+dataloader = load_retrieval_augmented_data(config, tokenizer, split_dataset=False)
+
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model = PythiaSeekForCausalLM.from_pretrained(config["model_name"], use_cache=False)
+model.to(device)
+model.eval()
+
+# Evaluate the model on the SQUAD dataset
+f1s = []
+exact_matches = []
+with torch.no_grad():
+    for batch in tqdm(dataloader):
+        outputs = model(input_ids=batch["input_ids"].to(device),
+                        labels=batch["labels"].to(device),
+                        encoder_hidden_states=batch["encoder_hidden_states"].to(device))
+
+        labels = batch["labels"]
+        mask = labels == -100
+
+        # since it's causal we predict the next token
+        predicted_tokens = outputs.logits.argmax(dim=-1)[:, :-1]
+        predicted_tokens[mask[:, 1:]] = tokenizer.pad_token_id
+        predicted = tokenizer.batch_decode(predicted_tokens, skip_special_tokens=True)
+
+        labels[mask] = tokenizer.pad_token_id
+        ground_truth = tokenizer.batch_decode(labels, skip_special_tokens=True)
+        print(f"Predicted: {predicted}")
+        print(f"Ground truth: {ground_truth}")
+
+        f1 = f1_score(predicted, ground_truth)
+        exact_match = exact_match_score(predicted, ground_truth)
+
+        f1s.extend(f1)
+        exact_matches.extend(exact_match)
+
+
+print(torch.tensor(f1s).mean())
+print(torch.tensor(exact_matches).to(torch.float32).mean())

gpt4all/eval/eval_squad_atlas_map.py

@@ -0,0 +1,116 @@
+import torch
+import torch.nn.functional as F
+from gpt4all.models import LetheForCausalLM
+from gpt4all.models.lethe.modeling_lethe import MemoryIndex
+from gpt4all.data.retrieval_dataloader import load_memory_augmented_data
+from gpt4all.train.metrics import f1_score, exact_match_score
+from gpt4all.utils.read import read_config
+from transformers import AutoTokenizer, AutoConfig
+from argparse import ArgumentParser
+from tqdm import tqdm
+from nomic import atlas
+from datasets import load_from_disk
+
+
+def calc_loss_per_item(logits, labels):
+    lm_logits = logits[:, :-1, :].contiguous()
+    lm_labels = labels[:, 1:].contiguous()
+    loss = F.cross_entropy(lm_logits.view(-1, lm_logits.size(-1)), lm_labels.view(-1), reduction="none")
+    loss = loss.reshape(labels.shape[0], -1).mean(dim=-1)
+
+    # return tensor of shape (B,) where B is the batch size
+    return loss.cpu().tolist()
+
+    
+def greedy_search(input_ids, model, tokenizer, max_new_tokens=100):
+    num_new_tokens = 0
+    with torch.no_grad():
+        while True:
+            if num_new_tokens >= max_new_tokens:
+                break
+            outputs = model(input_ids, save_kv=False)
+
+            new_tokens = torch.argmax(outputs.logits[:, -1, :], dim=-1)
+
+            input_ids = torch.cat([input_ids, new_tokens.unsqueeze(1)], dim=-1)
+            num_new_tokens += 1
+
+            if torch.equal(input_ids[0, -1].cpu(), torch.tensor(tokenizer.eos_token_id)):
+                break
+        
+    print(tokenizer.batch_decode(input_ids, skip_special_tokens=True))
+
+    return input_ids
+    
+
+parser = ArgumentParser()
+parser.add_argument("--config", type=str, default="config.yaml")
+
+args = parser.parse_args()
+
+config = read_config(args.config)
+
+tokenizer = AutoTokenizer.from_pretrained(config["tokenizer_name"], model_max_length=config["max_length"])
+if tokenizer.pad_token is None:
+    tokenizer.pad_token = tokenizer.eos_token
+
+dataloader = load_memory_augmented_data(config, tokenizer, split_dataset=False)
+
+dataset = load_from_disk(config["dataset_path"])
+
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model_config = AutoConfig.from_pretrained(config["model_name"])
+
+head_size = model_config.hidden_size // model_config.num_attention_heads
+index = MemoryIndex(head_size,
+            config["num_memories_per_index"],
+            model_config.num_attention_heads
+)
+model = LetheForCausalLM.from_pretrained(config["model_name"], 
+                                    revision=config['version'] if 'version' in config else None,
+                                    memory_attn_layer=config["memory_attn_layer"],
+                                    num_neighbors_to_retrieve=config["num_neighbors_to_retrieve"],
+                                    index=index,
+                                ).to(device)
+model.eval()
+
+# Evaluate the model on the SQUAD dataset
+losses = []
+with torch.no_grad():
+    for i, batch in enumerate(tqdm(dataloader)):
+        memories = batch["retrieved_context"]
+        memories = memories[:, :config["num_neighbors_to_store"], :]
+        memories = memories.reshape(-1, memories.shape[-1])
+
+        # need to set to eval so we don't do mem attn as it's slow
+        with torch.no_grad():
+            for chunk_start in range(0, memories.shape[0], config["mem_chunk_size"]):
+                chunk_end = min(memories.shape[0], chunk_start + config["mem_chunk_size"])
+                mem_chunk = memories[chunk_start:chunk_end]
+                model(input_ids=mem_chunk.to(device))
+
+        del memories
+        torch.cuda.empty_cache()
+        qa_inputs = batch["input_ids"]
+        qa_labels = batch["labels"]
+        for i in range(qa_inputs.shape[0]):
+            inputs = qa_inputs[i].to(device)
+            labels = qa_labels[i].to(device)
+            cutoff = torch.argmax((labels != -100).type(torch.float32))
+            greedy_search(inputs[:cutoff.item()].unsqueeze(0).to(device), model, tokenizer)
+            print(tokenizer.decode(inputs, skip_special_tokens=True))
+
+
+        # batch_loss = calc_loss_per_item(outputs.logits, qa_labels.to(device))
+        # losses.extend(batch_loss)
+        index.reset()
+
+        
+        
+dataset = dataset.add_column("loss", losses)
+
+dataset.save_to_disk("eval_squad_atlas_map")
+
+
+        

gpt4all/index/README.md

@@ -0,0 +1,34 @@
+# How to Tokenize and Embed
+
+Split text into chunks
+```
+python tokenize_texts.py
+```
+
+Embbed Texts
+
+```
+torchrun --master_port=29085 --nproc-per-node 8  embed_texts.py --ds_path=tokenized --batch_size=2048
+```
+
+
+Combine Embeddings and Build Index
+```
+python build_index.py --ds_path=wiki_sample_tokenized --embed_folder=wiki_sample_embedded
+```
+
+To use the Index
+
+```
+import hnswlib
+
+index = hnswlib.Index(space='l2', dim=384)
+index.load_index(<path to index>)
+```
+
+
+Prep index for train
+
+```
+CUDA_VISIBLE_DEVICES=7 torchrun --master_port=29086 --nproc-per-node 1 prep_index_for_train.py --config=../../configs/train/finetune_gptjr.yaml
+```

gpt4all/index/build_index.py

@@ -0,0 +1,102 @@
+import os
+from datasets import Dataset, concatenate_datasets
+import glob
+from argparse import ArgumentParser
+import hnswlib
+import pyarrow as pa
+import pyarrow.compute as pc
+from tqdm import tqdm
+
+
+def parse_args():
+    parser = ArgumentParser()
+    parser.add_argument("--ds_path", type=str, required=True)
+    parser.add_argument("--embed_folder", type=str, required=True)
+    parser.add_argument("--index_path", type=str, default="wiki-index")
+
+    return parser.parse_args()
+
+
+def concat_embedded(folder):
+    files = glob.glob(f"{folder}/*")
+
+    all_embeddings = []
+    for file in files:
+        all_embeddings.append(Dataset.load_from_disk(file))
+
+    all_embeddings = concatenate_datasets(all_embeddings)
+
+    return all_embeddings
+
+
+def join(original_ds, embedded_ds):
+    embedded_ds = embedded_ds.add_column("index", range(len(embedded_ds)))
+    embed_table = embedded_ds.data.table
+
+    seen = set()
+    indices = []
+    for i, id in enumerate(original_ds["id"]):
+        if id not in seen:
+            indices.append(i)
+            seen.add(id)
+
+    mask = pc.is_in(embed_table["index"], value_set=pa.array(indices, pa.int32()))
+    filtered_table = embed_table.filter(mask)
+
+    import pdb; pdb.set_trace()
+    # sort to make sure we're adding in right order
+    filtered_table = filtered_table.sort_by("id")
+
+    original_table = original_ds.data.table
+    original_table = original_table.sort_by("id")
+
+    original_table = original_table.append_column(
+        "embedding", filtered_table["embedding"]
+    )
+    # there's definitely a better way to do this but
+    # Dataset(original_table) throws `KeyError: 'embedding'`
+    joined = Dataset.from_dict(original_table.to_pydict())
+
+    return joined
+
+
+def build_index(orig_path, embed_folder_path, index_path):
+    if not os.path.exists(orig_path + "_embedded_with_text"):
+        # TODO: this doesn't work for large datasets!
+        # just convert to pandas and do this manually
+        ds = Dataset.load_from_disk(orig_path)
+        embed_ds = concat_embedded(embed_folder_path)
+        print("Concatenated embeddings")
+        print(f"Length: {len(ds)}")
+        print(f"Length: {len(embed_ds)}")
+        ds = join(ds, embed_ds)
+        ds = ds.add_column("index", range(len(ds)))
+        print("Saving to disk")
+        ds.save_to_disk(f"{orig_path}_embedded_with_text")
+    else:
+        ds = Dataset.load_from_disk(orig_path + "_embedded_with_text")
+
+    print(f"Length of ds: {len(ds)}")
+
+    print("Building index")
+    index = hnswlib.Index(space="cosine", dim=384)
+    # not sure what we should set M and ef_construction to
+    index.init_index(max_elements=len(ds), M=64, ef_construction=200)
+    print("Adding items")
+    chunk_size = 50_000
+    num_chunks = len(ds) // chunk_size
+    progbar = tqdm(total=num_chunks)
+    start = 0
+    while start < len(ds):
+        chunk = ds[start:start + chunk_size]
+        index.add_items(chunk["embedding"], chunk["id"], num_threads=64)
+        progbar.update(1)
+        start += chunk_size
+
+    print("Saving index")
+    index.save_index(index_path + ".bin")
+
+
+if __name__ == "__main__":
+    args = parse_args()
+    build_index(args.ds_path, args.embed_folder, args.index_path)

gpt4all/index/embed.py

@@ -0,0 +1,70 @@
+import torch
+from transformers import AutoTokenizer, AutoModel
+import torch.nn.functional as F
+
+
+class Embedder:
+    def __init__(self, model_name="sentence-transformers/all-MiniLM-L6-v2"):
+        self.tokenizer = AutoTokenizer.from_pretrained(model_name)
+        self.embedder = AutoModel.from_pretrained(model_name)
+        # hack
+        self.offset = self.tokenizer.model_max_length // 2
+
+    def _mean_pool(self, model_output, attention_mask):
+        token_embeddings = model_output[
+            0
+        ]  # First element of model_output contains all token embeddings
+        input_mask_expanded = (
+            attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
+        )
+        sentence_embeddings = torch.sum(
+            token_embeddings * input_mask_expanded, 1
+        ) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
+        return F.normalize(sentence_embeddings, p=2, dim=1)
+
+    def chunk_text(self, text):
+        tokenized_text = {"input_ids": [], "attention_mask": []}
+        tokenized = self.tokenizer(text)
+        tokenized_len = len(tokenized["input_ids"])
+        max_len = self.tokenizer.model_max_length
+        if tokenized_len > max_len:
+            start = 0
+            while start < tokenized_len:
+                tokenized_text["input_ids"].append(
+                    tokenized["input_ids"][start : start + max_len]
+                )
+                tokenized_text["attention_mask"].append(
+                    tokenized["attention_mask"][start : start + max_len]
+                )
+                # this could probably be done better
+                start += self.offset
+
+        else:
+            tokenized_text["input_ids"].append(tokenized["input_ids"])
+            tokenized_text["attention_mask"].append(tokenized["attention_mask"])
+
+        return tokenized_text
+
+    def tokenize(self, text):
+        return self.tokenizer(text, return_tensors="pt", truncation=True, padding="max_length")
+
+    def __call__(self, batch):
+        if isinstance(batch, str):
+            tokenized = self.tokenizer(batch, return_tensors="pt", truncation=True)
+            return self._mean_pool(
+                self.embedder(
+                    input_ids=tokenized["input_ids"],
+                    attention_mask=tokenized["attention_mask"],
+                ),
+                tokenized["attention_mask"],
+            )
+        else:
+            outputs = self.embedder(
+                input_ids=batch["input_ids"], attention_mask=batch["attention_mask"]
+            )
+            embedding = self._mean_pool(outputs, batch["attention_mask"])
+
+            return {"id": batch["id"], "embedding": embedding}
+
+    def to(self, device):
+        self.embedder.to(device)

gpt4all/index/embed_texts.py

@@ -0,0 +1,116 @@
+import os
+import torch.distributed as dist
+from argparse import ArgumentParser
+from datasets import Dataset
+from gpt4all.index.embed import Embedder
+from gpt4all.utils.distributed_utils import rank0_print
+from torch.utils.data import DataLoader, DistributedSampler
+from transformers.trainer_pt_utils import nested_numpify
+from transformers import BatchEncoding
+from tqdm import tqdm
+import numpy as np
+import torch
+from datasets import load_dataset
+
+
+# this isn't used but keeping in case we need it in the future
+# collate and pad inputs to the right shape
+class PadCollateInputs:
+    def __init__(self, tokenizer):
+        self.tokenizer = tokenizer
+
+    def __call__(self, batch):
+        mapped_inputs = {"input_ids": [], "attention_mask": []}
+        mapped_inputs["input_ids"] = [b["tokenized_chunk"] for b in batch]
+        mapped_inputs["attention_mask"] = [b["tokenized_attn_mask"] for b in batch]
+        encoding = BatchEncoding(mapped_inputs)
+
+        padded_inputs = self.tokenizer.pad(
+            encoding, padding="max_length", return_tensors="pt"
+        )
+        padded_inputs["id"] = [b["id"] for b in batch]
+
+        return padded_inputs
+
+
+def embed_texts(ds_path, batch_size, embed_on='text', save_to_disk=False, split='train'):
+    world_size = dist.get_world_size() if dist.is_initialized() else 1
+    rank0_print(f"World size: {world_size}")
+
+    if os.path.exists(ds_path):
+        dataset = Dataset.load_from_disk(ds_path)
+    else:
+        dataset = load_dataset(ds_path, split=split)
+    rank0_print(f"Dataset size: {len(dataset)}")
+
+    model = Embedder()
+
+    if "input_ids" not in dataset.column_names:
+        dataset = dataset.map(lambda x: model.tokenize(x[embed_on]), batched=True, num_proc=64)
+
+
+    columns_to_keep = ["id", "input_ids", "attention_mask"]
+    to_remove = [e for e in dataset.column_names if not e in columns_to_keep]
+    dataset = dataset.remove_columns(to_remove)
+
+    dataset = dataset.with_format("torch")
+
+    num_processes = dist.get_world_size() if dist.is_initialized() else 1
+    local_rank = dist.get_rank() if dist.is_initialized() else 0 
+
+
+    if num_processes > 1: 
+        sampler = DistributedSampler(
+            dataset,
+            shuffle=False,
+            drop_last=False,
+            num_replicas=num_processes,
+            rank=local_rank,
+        )
+    else:
+        sampler = None
+
+    dataloader = DataLoader(
+        dataset,
+        batch_size=batch_size,
+        sampler=sampler,
+        drop_last=False,
+    )
+
+    model.to(f"cuda:{local_rank}")
+    with torch.no_grad():
+        embedded_outputs = {"id": [], "embedding": []}
+        for batch in tqdm(dataloader, disable=local_rank != 0):
+            batch["input_ids"] = batch["input_ids"].to(f"cuda:{local_rank}")
+            batch["attention_mask"] = batch["attention_mask"].to(f"cuda:{local_rank}")
+            outputs = model(batch)
+            embedded_outputs["id"].extend(batch["id"])
+            embedded_outputs["embedding"].extend(outputs["embedding"])
+
+        embedded_outputs["embedding"] = nested_numpify(embedded_outputs["embedding"])
+        embedded_outputs["id"] = np.stack(embedded_outputs["id"])
+        embedded_outputs["embedding"] = np.stack(embedded_outputs["embedding"])
+
+    ds = Dataset.from_dict(embedded_outputs)
+
+    # feeling lazy, don't want to wait for all_gather to finish
+    # will load and concat in a single process in another script
+    if save_to_disk:
+        ds.save_to_disk(f"{ds_path}_embedded/{ds_path}_embedded_rank_{local_rank}")
+    return ds
+
+
+def main():
+    dist.init_process_group("nccl")
+    parser = ArgumentParser()
+    parser.add_argument("--ds_path", type=str, default="tokenized")
+    parser.add_argument("--batch_size", type=int, default=1)
+
+    args = parser.parse_args()
+
+    embed_texts(args.ds_path, args.batch_size, save_to_disk=True)
+
+
+if __name__ == "__main__":
+    # parse arguments by reading in a config
+    main()

gpt4all/index/prep_index_for_train.py

@@ -0,0 +1,94 @@
+import os
+import hnswlib
+import numpy as np
+import pyarrow as pa
+from datasets import Dataset
+import torch.distributed as dist
+from datasets import load_dataset
+from pyarrow import compute as pc
+from argparse import ArgumentParser
+from gpt4all.utils.read import read_config
+from gpt4all.index.embed_texts import embed_texts
+from tqdm import tqdm
+
+CHUNK_SIZE = 2048
+
+def parse_args():
+    parser = ArgumentParser()
+    parser.add_argument("--config", type=str, default="config.yaml")
+    parser.add_argument("--split", type=str, default="train")
+    parser.add_argument("--k", type=int, default=5)
+
+    return parser.parse_args()
+
+    
+def prep_index():
+    args = parse_args()
+    config = read_config(args.config)
+    index = hnswlib.Index(space=config['index_space'], dim=config['index_dim'])
+    print("loading index")
+    index.load_index(config['index_path'])
+
+    # load query dataset
+    ds_path = config['dataset_path']
+
+    # load retrieval dataset
+    print("loading retrieval dataset")
+    print(config["index_database"])
+    if os.path.exists(config['index_database']):
+        retrieval_dataset = Dataset.load_from_disk(config['index_database'])
+    else:
+        retrieval_dataset = load_dataset(config['index_database'], split="train")
+
+    # vectorize queries
+    query_vector_path =  f"{ds_path}_queries_embedded/{ds_path}_embedded_{args.split}"
+    if not os.path.exists(query_vector_path):
+        print('Embedding dataset...')
+        ds = embed_texts(ds_path,
+                         config['batch_size'],
+                         embed_on=config['query_embedding_field'],
+                         save_to_disk=False,
+                         split=args.split)
+        ds.save_to_disk(query_vector_path)
+    else:
+        print('Found cached embedding dataset!')
+        ds = Dataset.load_from_disk(query_vector_path)
+
+    #build training dataset
+    train_dataset = load_dataset(ds_path, split=args.split)
+
+    # search the index for each query
+    neighbor_text_column = []
+    neighbor_ids_column = []
+    for _, chunk_start in enumerate(tqdm(range(0, len(ds), CHUNK_SIZE))):
+        chunk_end = chunk_start + CHUNK_SIZE
+        chunk = ds[chunk_start:chunk_end]
+        query_vectors = np.array(chunk['embedding'])
+        neighbor_ids, _ = index.knn_query(query_vectors, k=args.k, num_threads=-1) # neighbor ids is of shape [n_queries, n_neighbors]
+        value_set = pa.array([str(e) for e in neighbor_ids.flatten()])
+        neighbor_objs = retrieval_dataset._data.filter(pc.is_in(retrieval_dataset._data['id'], value_set))
+
+        # build mapping between indices and embeddings
+        neighbor_id_list = neighbor_objs['id']
+        documents = neighbor_objs['text']
+        idx_to_embedding = {idx.as_py(): documents[i] for i, idx in enumerate(neighbor_id_list)}
+
+        neighbor_text = []
+        for cur_neighbor_ids in neighbor_ids:
+            cur_embs = [idx_to_embedding[id].as_py() for id in cur_neighbor_ids]
+            neighbor_text.append(cur_embs)
+
+        neighbor_text_column.extend(neighbor_text)
+        neighbor_ids_column.extend(neighbor_ids)
+
+    print("adding neighbor ids")
+    train_dataset = train_dataset.add_column('neighbor_ids', neighbor_ids_column)
+    print("adding neighbors")
+    train_dataset = train_dataset.add_column('neighbor_text', neighbor_text_column)
+
+    supplemented_dataset_path = f"{ds_path}_supplemented_{args.split}/"
+    train_dataset.save_to_disk(supplemented_dataset_path)
+
+
+if __name__ == "__main__":
+    prep_index()

gpt4all/index/tokenize_texts.py

@@ -0,0 +1,72 @@
+from datasets import load_dataset
+from argparse import ArgumentParser
+from gpt4all.index.embed import Embedder
+
+
+def parse_args():
+    parser = ArgumentParser()
+    # fmt: off
+    parser.add_argument("--tokenized_save_path", type=str, default="tokenized")
+    parser.add_argument("--ds_name", type=str, default="wikipedia")
+    parser.add_argument("--ds_version", type=str, default="20220301.simple")
+    parser.add_argument("--sbert_model", type=str, default="sentence-transformers/all-MiniLM-L6-v2")
+    # fmt: on
+
+    return parser.parse_args()
+
+
+def tokenize_texts(examples, embedder):
+    split_data = {k: [] for k in examples.keys()}
+    split_data["tokenized_chunk"] = []
+    split_data["tokenized_attn_mask"] = []
+
+    keys = [k for k in examples.keys() if k != "text"]
+    for i, text in enumerate(examples["text"]):
+        tokenized_text = embedder.chunk_text(text)
+        # do we want to add sep/cls tokens to beginning and end?
+        decoded_text = embedder.tokenizer.batch_decode(
+            sequences=tokenized_text["input_ids"]
+        )
+
+        num_splits = len(tokenized_text["input_ids"])
+        split_data["id"].extend(
+            [f"{examples['id'][i]}_split_{j}" for j in range(num_splits)]
+        )
+
+        for col in keys:
+            if col != "id":
+                split_data[col].extend(
+                    [examples[col][i]] * len(tokenized_text["input_ids"])
+                )
+
+        split_data["text"].extend(decoded_text)
+        split_data["tokenized_chunk"].extend(tokenized_text["input_ids"])
+        split_data["tokenized_attn_mask"].extend(tokenized_text["attention_mask"])
+
+    return split_data
+
+
+def chunk_dataset(
+    ds_name="wikipedia",
+    version="20220301.simple",
+    sbert_model="sentence-transformers/all-MiniLM-L6-v2",
+    save_path="tokenized",
+):
+    dataset = load_dataset(ds_name, version, split="train")
+    print(len(dataset))
+    embedder = Embedder(sbert_model)
+    dataset = dataset.map(
+        lambda x: tokenize_texts(x, embedder), batched=True, num_proc=64
+    )
+
+    dataset.save_to_disk(save_path)
+
+
+if __name__ == "__main__":
+    args = parse_args()
+    chunked_dataset = chunk_dataset(
+        ds_name=args.ds_name,
+        version=args.ds_version,
+        sbert_model=args.sbert_model,
+        save_path=args.tokenized_save_path,
+    )

gpt4all/inference/combine_synth_data.py

@@ -0,0 +1,57 @@
+import glob
+from argparse import ArgumentParser
+from datasets import Dataset, load_from_disk, concatenate_datasets
+
+
+PROMPT = "Write a question answer pair based on the following context. If the context isn't specific enough, ignore and return 'No question answer pair`. Context : {}\n"
+
+def load_synth_data(data_dir):
+    files = glob.glob(data_dir + "/*")
+
+    ds = concatenate_datasets([load_from_disk(f) for f in files])
+    table = ds.data.table
+
+    filtered = table.filter(table["valid"])
+    ds = Dataset.from_dict(filtered.to_pydict())
+    return ds 
+
+    
+def remove_prompt(examples):
+    outputs = {"text": [], "generated": [], "question": [], "answer": []}
+    for context, generated in zip(examples["text"], examples["generated"]):
+        prompt_w_ctx = PROMPT.format(context)
+        gen_wo_ctx = generated[len(prompt_w_ctx):]
+
+        assert prompt_w_ctx not in gen_wo_ctx
+
+        question = gen_wo_ctx.split("Answer:")[0].replace("Question:", "").strip()
+        answer = gen_wo_ctx.split("Answer:")[1].strip()
+
+        outputs["text"].append(context)
+        outputs["generated"].append(gen_wo_ctx)
+        outputs["question"].append(question)
+        outputs["answer"].append(answer)
+
+    return outputs
+
+        
+
+
+def combine_synth_data(data_dir):
+    ds = load_synth_data(data_dir)
+
+    ds = ds.map(lambda ele: remove_prompt(ele), batched=True, num_proc=64)
+
+    ds.save_to_disk(f"synth_data_combined_{len(ds)/1000:.0f}")
+
+
+
+if __name__ == "__main__":
+    parser = ArgumentParser()
+    parser.add_argument("--dataset_dir", type=str, required=True)
+
+    args = parser.parse_args()
+
+    combine_synth_data(args.dataset_dir)
+
+    

gpt4all/inference/generate.py

@@ -1,6 +1,6 @@
 from transformers import AutoModelForCausalLM, AutoTokenizer
 from peft import PeftModelForCausalLM
-from read import read_config
+from gpt4all.utils.read import read_config
 from argparse import ArgumentParser
 import torch
 import time

gpt4all/inference/generate_synth_data.py

@@ -0,0 +1,149 @@
+from argparse import ArgumentParser
+from datasets import load_dataset, Dataset
+import torch
+from torch.utils.data import DataLoader, DistributedSampler
+from accelerate.utils import set_seed
+from transformers import AutoTokenizer, AutoModelForCausalLM, DefaultDataCollator 
+from gpt4all.utils.read import read_config
+from gpt4all.utils.distributed_utils import rank0_print, main_process_first
+from tqdm import tqdm
+import pyarrow.compute as pc
+import pyarrow as pa
+import torch.distributed as dist
+
+
+PROMPT = "Write a question answer pair based on the following context. If the context isn't specific enough, ignore and return 'No question answer pair`. Context : {}\n"
+
+
+def prepare_data(config, tokenizer, num_processes, local_rank):
+    dataset = load_dataset(config["dataset_path"], split="train")
+    dataset = dataset.remove_columns("embedding")
+
+    shuffled = dataset.shuffle(seed=config["seed"])
+    indices = shuffled[:config["max_generations"]]["id"]
+
+    table = dataset.data
+    mask = pc.is_in(table["id"], value_set=pa.array(indices, pa.int32()))
+    filtered_table = table.filter(mask)
+
+    # convert from pyarrow to Dataset
+    orig_dataset = Dataset.from_dict(filtered_table.to_pydict())
+
+    dataset = orig_dataset.map(lambda ele: {"prompted_text": [PROMPT.format(context) for context in ele["text"]]},
+                          batched=True,
+                          num_proc=config["num_proc"] if "num_proc" in config else None)
+
+    dataset = dataset.map(lambda ele: {"prompt_len": [len(prompt) for prompt in ele["prompted_text"]]}, batched=True,
+                          num_proc=config["num_proc"] if "num_proc" in config else None)
+
+    dataset = dataset.sort("prompt_len")
+    dataset = dataset.map(lambda ele: tokenizer(ele["prompted_text"], return_tensors="pt", padding="longest", truncation=True,
+                                                max_length=tokenizer.model_max_length - config["max_new_tokens"]), batched=True,
+                                                batch_size=num_processes * config["batch_size"],
+                         )
+
+    columns_to_keep = ["id", "input_ids", "attention_mask"]
+    col_names_to_rm = [col for col in dataset.column_names if col not in columns_to_keep]
+
+    dataset = dataset.remove_columns(col_names_to_rm)
+
+    sampler = DistributedSampler(
+        dataset,
+        shuffle=False,
+        drop_last=True,
+        num_replicas=num_processes,
+        rank=local_rank,
+    )
+
+    dataloader = DataLoader(
+            dataset,
+            batch_size=config["batch_size"],
+            collate_fn=DefaultDataCollator(),
+            sampler=sampler,
+            drop_last=True
+        )
+
+    return dataloader, orig_dataset
+
+def generate_data(config):
+    set_seed(config["seed"])
+
+    rank0_print(f"World size: {dist.get_world_size()}")
+
+    tokenizer = AutoTokenizer.from_pretrained(config["tokenizer_name"])
+    # since we're doing generation, pad left for autoregressive generation
+    tokenizer.padding_side = "left"
+    if tokenizer.pad_token is None:
+        tokenizer.pad_token = tokenizer.eos_token
+
+    num_processes = dist.get_world_size()
+    local_rank = dist.get_rank()
+
+    dataloader, dataset = prepare_data(config, tokenizer, num_processes, local_rank)
+
+    dist.barrier()
+    print(dataset[:10]["id"])
+        
+    model = AutoModelForCausalLM.from_pretrained(config["model_name"],
+                                                       revision=config["revision"] if "revision" in config else None,
+                                                       use_cache=True,
+                                                       torch_dtype=torch.bfloat16,)
+    model.to(f"cuda:{local_rank}")
+
+    synth_data = []
+    valid = []
+    ids = []
+    total_valid = 0
+
+    with torch.no_grad():
+        for i, batch in enumerate(tqdm(dataloader, disable=local_rank != 0)):
+            # keep this simple for now, can add temperature and other sampling techniques later
+            generated = model.generate(batch["input_ids"].to(model.device),
+                                       attention_mask=batch["attention_mask"].to(model.device),
+                                       max_new_tokens=config["max_new_tokens"])
+
+            decoded = tokenizer.batch_decode(generated, skip_special_tokens=True)
+            num_valid = ["\nQuestion:" in t and "\nAnswer:" in t for t in decoded]
+            rank0_print(f"Num valid: {sum(num_valid)/ len(num_valid):.2f}")
+            total_valid += sum(num_valid)
+
+            synth_data.extend(decoded)
+            valid.extend(num_valid)
+            ids.extend(batch["id"].tolist())
+
+            if i > 0 and i % config["save_every"] == 0:
+                table = dataset.data.table
+                mask = pc.is_in(table["id"], value_set=pa.array(ids, pa.int32()))
+                filtered_table = table.filter(mask)
+
+                chunk_table = pa.Table.from_pydict({"id": ids, "generated": synth_data, "valid": valid})
+                joined = filtered_table.join(chunk_table, "id")
+                curr_dataset = Dataset.from_dict(joined.to_pydict())
+                curr_dataset.save_to_disk(f'{config["output_path"]}/chunk_{i}_rank_{local_rank}')
+
+    table = dataset.data.table
+    mask = pc.is_in(table["id"], value_set=pa.array(ids, pa.int32()))
+    filtered_table = table.filter(mask)
+
+    chunk_table = pa.Table.from_pydict({"id": ids, "generated": synth_data, "valid": valid})
+    joined = filtered_table.join(chunk_table, "id")
+    full_dataset = Dataset.from_dict(joined.to_pydict())
+    full_dataset.save_to_disk(f'{config["output_path"]}_{config["max_generations"]}_rank_{local_rank}')
+
+    rank0_print(f"Total valid: {total_valid}/{config['max_generations']}")
+
+    
+def main():
+    dist.init_process_group("nccl")
+    parser = ArgumentParser()
+    parser.add_argument("--config", type=str, default="config.yaml")
+
+    args = parser.parse_args()
+    config = read_config(args.config)
+
+    generate_data(config)
+
+
+if __name__ == "__main__":
+    # parse arguments by reading in a config
+    main()

gpt4all/inference/inference.py

@@ -2,13 +2,14 @@ from transformers import AutoModelForCausalLM, AutoTokenizer
 import torch
 import torch.nn as nn
 from argparse import ArgumentParser
-from read import read_config
-from accelerate.utils import  set_seed
-from data import load_data_for_inference
+from gpt4all.utils.read import read_config
+from accelerate.utils import set_seed
+from gpt4all.data.instruction_tuning_dataloader import load_data_for_inference
+from gpt4all.utils.distributed_utils import rank0_print
 from tqdm import tqdm
-from datasets import  Dataset
+from datasets import Dataset
 import torch.distributed as dist
-from transformers.trainer_pt_utils import  nested_numpify
+from transformers.trainer_pt_utils import nested_numpify
 from transformers import DefaultDataCollator
 from torch.utils.data import DataLoader, DistributedSampler
 import numpy as np
@@ -21,56 +22,64 @@ def calc_cross_entropy_no_reduction(lm_logits, labels):
     shift_logits = lm_logits[..., :-1, :].contiguous()
     shift_labels = labels[..., 1:].contiguous()
     # Flatten the tokens
-    loss_fct = nn.CrossEntropyLoss(reduction='none')
+    loss_fct = nn.CrossEntropyLoss(reduction="none")
     loss = loss_fct(shift_logits.permute(0, 2, 1), shift_labels).mean(dim=1)
 
     return loss
 
 
-def rank0_print(msg):
-    if dist.get_rank() == 0:
-        print(msg)
-        
-
 def inference(config):
-    set_seed(config['seed'])
+    set_seed(config["seed"])
 
     rank0_print(f"World size: {dist.get_world_size()}")
 
-    tokenizer = AutoTokenizer.from_pretrained(config['tokenizer_name'], model_max_length=config['max_length'])
+    tokenizer = AutoTokenizer.from_pretrained(
+        config["tokenizer_name"], model_max_length=config["max_length"]
+    )
     # llama has no pad token, set it to new token
     if tokenizer.pad_token is None:
         tokenizer.pad_token = tokenizer.eos_token
 
-        
-    train_dataset, val_dataset = load_data_for_inference(config, tokenizer) 
+    train_dataset, val_dataset = load_data_for_inference(config, tokenizer)
 
     num_processes = dist.get_world_size()
     local_rank = dist.get_rank()
 
-    train_sampler = DistributedSampler(train_dataset, shuffle=False, drop_last=True, num_replicas=num_processes, rank=local_rank)
+    train_sampler = DistributedSampler(
+        train_dataset,
+        shuffle=False,
+        drop_last=True,
+        num_replicas=num_processes,
+        rank=local_rank,
+    )
     train_dataloader = DataLoader(
         train_dataset,
         collate_fn=DefaultDataCollator(),
         batch_size=config["batch_size"],
         sampler=train_sampler,
-        drop_last=True
+        drop_last=True,
     )
 
-    val_sampler = DistributedSampler(val_dataset, shuffle=False, drop_last=True, num_replicas=num_processes, rank=local_rank)
+    val_sampler = DistributedSampler(
+        val_dataset,
+        shuffle=False,
+        drop_last=True,
+        num_replicas=num_processes,
+        rank=local_rank,
+    )
     val_dataloader = DataLoader(
         val_dataset,
         collate_fn=DefaultDataCollator(),
         batch_size=config["batch_size"],
         sampler=val_sampler,
-        drop_last=True
+        drop_last=True,
     )
 
-
-    model = AutoModelForCausalLM.from_pretrained(config["model_name"], 
-                                                    trust_remote_code=True,
-                                                    torch_dtype=torch.bfloat16,
-                                                    ) 
+    model = AutoModelForCausalLM.from_pretrained(
+        config["model_name"],
+        trust_remote_code=True,
+        torch_dtype=torch.bfloat16,
+    )
     model.to(f"cuda:{local_rank}")
 
     with torch.no_grad():
@@ -78,14 +87,18 @@ def inference(config):
         for batch in tqdm(train_dataloader, disable=local_rank != 0):
             batch["input_ids"] = batch["input_ids"].to(f"cuda:{local_rank}")
             batch["labels"] = batch["labels"].to(f"cuda:{local_rank}")
-            outputs = model(input_ids=batch["input_ids"], labels=batch["labels"], output_hidden_states=True)
+            outputs = model(
+                input_ids=batch["input_ids"],
+                labels=batch["labels"],
+                output_hidden_states=True,
+            )
             loss = calc_cross_entropy_no_reduction(outputs.logits, batch["labels"])
             train_outputs["loss"].extend(loss)
 
             embeddings = outputs.hidden_states[-1]
             batch_size = batch["input_ids"].shape[0]
             sequence_lengths = []
-            # since we use mutiturn with multiple <|endoftext|>, we need to find the place where 
+            # since we use mutiturn with multiple <|endoftext|>, we need to find the place where
             # <|endoftext|> is repeated
             for item in batch["input_ids"]:
                 indices = torch.where(item == tokenizer.pad_token_id)[0]
@@ -101,7 +114,9 @@ def inference(config):
                     sequence_lengths.append(len(item) - 1)
 
             sequence_lengths = torch.tensor(sequence_lengths)
-            pooled_logits = embeddings[torch.arange(batch_size, device=embeddings.device), sequence_lengths]
+            pooled_logits = embeddings[
+                torch.arange(batch_size, device=embeddings.device), sequence_lengths
+            ]
 
             train_outputs["embeddings"].append(pooled_logits)
             train_outputs["index"].extend(batch["index"].to(model.device))
@@ -120,29 +135,40 @@ def inference(config):
         # compute mask in pyarrow since it's super fast
         # ty @bmschmidt for showing me this!
         table = train_dataset.data
-        mask = pc.is_in(table['index'], value_set=pa.array(curr_idx, pa.int32()))
+        mask = pc.is_in(table["index"], value_set=pa.array(curr_idx, pa.int32()))
         filtered_table = table.filter(mask)
         # convert from pyarrow to Dataset
         filtered_train = Dataset.from_dict(filtered_table.to_pydict())
 
         filtered_train = filtered_train.add_column("embeddings", df_train["embeddings"])
         filtered_train = filtered_train.add_column("loss", df_train["loss"])
-        filtered_train = filtered_train.add_column("is_train", [True] * len(filtered_train))
+        filtered_train = filtered_train.add_column(
+            "is_train", [True] * len(filtered_train)
+        )
 
-        filtered_train.to_json(f"inference/epoch_2_embeddings_train_shard_{local_rank}.jsonl", lines=True, orient="records", num_proc=64)
+        filtered_train.to_json(
+            f"inference/epoch_2_embeddings_train_shard_{local_rank}.jsonl",
+            lines=True,
+            orient="records",
+            num_proc=64,
+        )
 
         val_outputs = {"loss": [], "embeddings": [], "index": []}
         for batch in tqdm(val_dataloader, disable=local_rank != 0):
             batch["input_ids"] = batch["input_ids"].to(f"cuda:{local_rank}")
             batch["labels"] = batch["labels"].to(f"cuda:{local_rank}")
-            outputs = model(input_ids=batch["input_ids"], labels=batch["labels"], output_hidden_states=True)
+            outputs = model(
+                input_ids=batch["input_ids"],
+                labels=batch["labels"],
+                output_hidden_states=True,
+            )
             loss = calc_cross_entropy_no_reduction(outputs.logits, batch["labels"])
             val_outputs["loss"].extend(loss)
 
             embeddings = outputs.hidden_states[-1]
             batch_size = batch["input_ids"].shape[0]
             sequence_lengths = []
-            # since we use mutiturn with multiple <|endoftext|>, we need to find the place where 
+            # since we use mutiturn with multiple <|endoftext|>, we need to find the place where
             # <|endoftext|> is repeated
             for item in batch["input_ids"]:
                 indices = torch.where(item == tokenizer.pad_token_id)[0]
@@ -158,7 +184,9 @@ def inference(config):
                     sequence_lengths.append(len(item) - 1)
 
             sequence_lengths = torch.tensor(sequence_lengths)
-            pooled_logits = embeddings[torch.arange(batch_size, device=embeddings.device), sequence_lengths]
+            pooled_logits = embeddings[
+                torch.arange(batch_size, device=embeddings.device), sequence_lengths
+            ]
 
             val_outputs["embeddings"].append(pooled_logits)
             val_outputs["index"].extend(batch["index"].to(model.device))
@@ -176,7 +204,7 @@ def inference(config):
         # compute mask in pyarrow since it's super fast
         # ty @bmschmidt for showing me this!
         table = val_dataset.data
-        mask = pc.is_in(table['index'], value_set=pa.array(curr_idx, pa.int32()))
+        mask = pc.is_in(table["index"], value_set=pa.array(curr_idx, pa.int32()))
         filtered_table = table.filter(mask)
         # convert from pyarrow to Dataset
         filtered_val = Dataset.from_dict(filtered_table.to_pydict())
@@ -184,8 +212,13 @@ def inference(config):
         filtered_val = filtered_val.add_column("loss", df_val["loss"])
         filtered_val = filtered_val.add_column("is_train", [False] * len(filtered_val))
 
-        filtered_val.to_json(f"inference/epoch_2_embeddings_val_shard_{local_rank}.jsonl", lines=True, orient="records", num_proc=64)
-    
+        filtered_val.to_json(
+            f"inference/epoch_2_embeddings_val_shard_{local_rank}.jsonl",
+            lines=True,
+            orient="records",
+            num_proc=64,
+        )
+
 
 def main():
     dist.init_process_group("nccl")
@@ -201,4 +234,3 @@ def main():
 if __name__ == "__main__":
     # parse arguments by reading in a config
     main()
-    

gpt4all/models/__init__.py

@@ -0,0 +1,18 @@
+from .gpt_jr.configuration_gpt_jr import GPTJRConfig
+from .gpt_jr.modeling_gpt_jr import GPTJRForCausalLM
+
+from .pythiaseek import PythiaSeekForCausalLM, PythiaSeekConfig
+from .pythia_retro import PythiaRetroForCausalLM, PythiaRetroConfig
+from .lethe import LetheConfig, LetheForCausalLM
+
+
+__all__ = [
+    "GPTJRConfig",
+    "GPTJRForCausalLM",
+    "PythiaSeekConfig",
+    "PythiaSeekForCausalLM",
+    "PythiaRetroConfig",
+    "PythiaRetroForCausalLM",
+    "LetheConfig",
+    "LetheForCausalLM"
+]

gpt4all/models/gpt_jr/configuration_gpt_jr.py

@@ -0,0 +1,151 @@
+# coding=utf-8
+# Copyright 2021 The EleutherAI and HuggingFace Teams. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+""" GPT-J model configuration"""
+from collections import OrderedDict
+from typing import Any, List, Mapping, Optional
+
+from transformers import PreTrainedTokenizer, TensorType, is_torch_available
+from transformers.configuration_utils import PretrainedConfig
+from transformers.onnx import OnnxConfigWithPast, PatchingSpec
+from transformers.utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+GPTJ_PRETRAINED_CONFIG_ARCHIVE_MAP = {
+    "EleutherAI/gpt-j-6B": "https://huggingface.co/EleutherAI/gpt-j-6B/resolve/main/config.json",
+    # See all GPT-J models at https://huggingface.co/models?filter=gpt_j
+}
+
+
+class GPTJRConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`GPTJModel`]. It is used to instantiate a GPT-J
+    model according to the specified arguments, defining the model architecture. Instantiating a configuration with the
+    defaults will yield a similar configuration to that of the GPT-J
+    [EleutherAI/gpt-j-6B](https://huggingface.co/EleutherAI/gpt-j-6B) architecture. Configuration objects inherit from
+    [`PretrainedConfig`] and can be used to control the model outputs. Read the documentation from [`PretrainedConfig`]
+    for more information.
+
+    Args:
+        vocab_size (`int`, *optional*, defaults to 50400):
+            Vocabulary size of the GPT-J model. Defines the number of different tokens that can be represented by the
+            `inputs_ids` passed when calling [`GPTJModel`].
+        n_positions (`int`, *optional*, defaults to 2048):
+            The maximum sequence length that this model might ever be used with. Typically set this to something large
+            just in case (e.g., 512 or 1024 or 2048).
+        n_embd (`int`, *optional*, defaults to 4096):
+            Dimensionality of the embeddings and hidden states.
+        n_layer (`int`, *optional*, defaults to 28):
+            Number of hidden layers in the Transformer encoder.
+        n_head (`int`, *optional*, defaults to 16):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        rotary_dim (`int`, *optional*, defaults to 64):
+            Number of dimensions in the embedding that Rotary Position Embedding is applied to.
+        n_inner (`int`, *optional*, defaults to None):
+            Dimensionality of the inner feed-forward layers. `None` will set it to 4 times n_embd
+        activation_function (`str`, *optional*, defaults to `"gelu_new"`):
+            Activation function, to be selected in the list `["relu", "silu", "gelu", "tanh", "gelu_new"]`.
+        resid_pdrop (`float`, *optional*, defaults to 0.1):
+            The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
+        embd_pdrop (`int`, *optional*, defaults to 0.1):
+            The dropout ratio for the embeddings.
+        attn_pdrop (`float`, *optional*, defaults to 0.1):
+            The dropout ratio for the attention.
+        layer_norm_epsilon (`float`, *optional*, defaults to 1e-5):
+            The epsilon to use in the layer normalization layers.
+        initializer_range (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        scale_attn_weights (`bool`, *optional*, defaults to `True`):
+            Scale attention weights by dividing by sqrt(hidden_size).
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should return the last key/values attentions (not used by all models).
+
+    Example:
+
+    ```python
+    >>> from transformers import GPTJModel, GPTJConfig
+
+    >>> # Initializing a GPT-J 6B configuration
+    >>> configuration = GPTJConfig()
+
+    >>> # Initializing a model from the configuration
+    >>> model = GPTJModel(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+    model_type = "gptj"
+    attribute_map = {
+        "max_position_embeddings": "n_positions",
+        "hidden_size": "n_embd",
+        "num_attention_heads": "n_head",
+        "num_hidden_layers": "n_layer",
+    }
+
+    def __init__(
+        self,
+        vocab_size=50400,
+        n_positions=2048,
+        n_embd=4096,
+        n_layer=28,
+        n_head=16,
+        rotary_dim=64,
+        n_inner=None,
+        activation_function="gelu_new",
+        resid_pdrop=0.0,
+        embd_pdrop=0.0,
+        attn_pdrop=0.0,
+        layer_norm_epsilon=1e-5,
+        initializer_range=0.02,
+        scale_attn_weights=True,
+        use_cache=True,
+        bos_token_id=50256,
+        eos_token_id=50256,
+        tie_word_embeddings=False,
+        encoder_dim=4096,
+        total_alpha_steps=0,
+        initial_alpha=1,
+        final_alpha=.5,
+        **kwargs
+    ):
+        self.vocab_size = vocab_size
+        self.n_positions = n_positions
+        self.n_embd = n_embd
+        self.n_layer = n_layer
+        self.n_head = n_head
+        self.n_inner = n_inner
+        self.rotary_dim = rotary_dim
+        self.activation_function = activation_function
+        self.resid_pdrop = resid_pdrop
+        self.embd_pdrop = embd_pdrop
+        self.attn_pdrop = attn_pdrop
+        self.layer_norm_epsilon = layer_norm_epsilon
+        self.initializer_range = initializer_range
+        self.scale_attn_weights = scale_attn_weights
+        self.use_cache = use_cache
+
+        self.bos_token_id = bos_token_id
+        self.eos_token_id = eos_token_id
+
+        self.encoder_dim = encoder_dim
+
+        self.total_alpha_steps = total_alpha_steps
+        self.initial_alpha = initial_alpha
+        self.final_alpha = final_alpha
+
+        super().__init__(
+            bos_token_id=bos_token_id, eos_token_id=eos_token_id, tie_word_embeddings=tie_word_embeddings, **kwargs
+        )

gpt4all/models/gpt_jr/modeling_gpt_jr.py

@@ -0,0 +1,953 @@
+# coding=utf-8
+# Copyright 2021 The EleutherAI and HuggingFace Teams. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+""" PyTorch GPT-J model."""
+
+from typing import Optional, Tuple, Union
+
+import math
+import torch
+import torch.utils.checkpoint
+from torch import nn
+from torch.nn import CrossEntropyLoss
+
+from transformers import AutoModel
+from transformers.activations import ACT2FN
+from transformers.modeling_outputs import (
+    BaseModelOutputWithPast,
+    CausalLMOutputWithPast,
+)
+from transformers.modeling_utils import PreTrainedModel
+from transformers.utils import logging
+from transformers.utils.model_parallel_utils import assert_device_map, get_device_map
+from gpt4all.models.gpt_jr.configuration_gpt_jr import GPTJRConfig
+
+
+logger = logging.get_logger(__name__)
+
+
+GPTJR_PRETRAINED_MODEL_ARCHIVE_LIST = [
+    "EleutherAI/gpt-j-6B",
+    # See all GPT-J models at https://huggingface.co/models?filter=gptj
+]
+
+
+def fixed_pos_embedding(x, seq_dim=1, seq_len=None):
+    dim = x.shape[-1]
+    if seq_len is None:
+        seq_len = x.shape[seq_dim]
+    inv_freq = 1.0 / (10000 ** (torch.arange(0, dim, 2) / dim))
+    sinusoid_inp = (
+        torch.einsum("i , j -> i j", torch.arange(seq_len, dtype=torch.float), inv_freq).to(x.device).float()
+    )
+    return torch.sin(sinusoid_inp), torch.cos(sinusoid_inp)
+
+
+def rotate_every_two(x):
+    x1 = x[:, :, :, ::2]
+    x2 = x[:, :, :, 1::2]
+    x = torch.stack((-x2, x1), dim=-1)
+    return x.flatten(-2)  # in einsum notation: rearrange(x, '... d j -> ... (d j)')
+
+
+def duplicate_interleave(m):
+    """
+    A simple version of `torch.repeat_interleave` for duplicating a matrix while interleaving the copy.
+    """
+    dim0 = m.shape[0]
+    m = m.view(-1, 1)  # flatten the matrix
+    m = m.repeat(1, 2)  # repeat all elements into the 2nd dimension
+    m = m.view(dim0, -1)  # reshape into a matrix, interleaving the copy
+    return m
+
+
+def apply_rotary_pos_emb(x, sincos, offset=0):
+    sin, cos = map(lambda t: duplicate_interleave(t)[None, offset : x.shape[1] + offset, None, :], sincos)
+    # einsum notation for lambda t: repeat(t[offset:x.shape[1]+offset,:], "n d -> () n () (d j)", j=2)
+    return (x * cos) + (rotate_every_two(x) * sin)
+
+    
+
+
+class GPTJRAttention(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+
+        max_positions = config.max_position_embeddings
+        self.register_buffer(
+            "bias",
+            torch.tril(torch.ones((max_positions, max_positions), dtype=torch.uint8)).view(
+                1, 1, max_positions, max_positions
+            ),
+        )
+        self.register_buffer("masked_bias", torch.tensor(-1e9))
+
+        self.attn_dropout = nn.Dropout(config.attn_pdrop)
+        self.resid_dropout = nn.Dropout(config.resid_pdrop)
+
+        self.embed_dim = config.hidden_size
+        self.num_attention_heads = config.num_attention_heads
+        self.head_dim = self.embed_dim // self.num_attention_heads
+        if self.head_dim * self.num_attention_heads != self.embed_dim:
+            raise ValueError(
+                f"embed_dim must be divisible by num_attention_heads (got `embed_dim`: {self.embed_dim} and"
+                f" `num_attention_heads`: {self.num_attention_heads})."
+            )
+        self.scale_attn = torch.sqrt(torch.tensor(self.head_dim, dtype=torch.float32)).to(torch.get_default_dtype())
+
+        self.k_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+        self.v_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+        self.q_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+        self.out_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+        self.rotary_dim = None
+        if config.rotary_dim is not None:
+            self.rotary_dim = config.rotary_dim
+
+    def _split_heads(self, tensor, num_attention_heads, attn_head_size, rotary):
+        """
+        Splits hidden dim into attn_head_size and num_attention_heads
+        """
+        new_shape = tensor.size()[:-1] + (num_attention_heads, attn_head_size)
+        tensor = tensor.view(new_shape)
+        if rotary:
+            return tensor
+        if len(tensor.shape) == 5:
+            return tensor.permute(0, 1, 3, 2, 4)  # (batch, blocks, head, block_length, head_features)
+        elif len(tensor.shape) == 4:
+            return tensor.permute(0, 2, 1, 3)  # (batch, head, seq_length, head_features)
+        else:
+            raise ValueError(f"Input tensor rank should be one of [4, 5], but is: {len(tensor.shape)}")
+
+    def _merge_heads(self, tensor, num_attention_heads, attn_head_size):
+        """
+        Merges attn_head_size dim and num_attn_heads dim into hidden dim
+        """
+        if len(tensor.shape) == 5:
+            tensor = tensor.permute(0, 1, 3, 2, 4).contiguous()
+        elif len(tensor.shape) == 4:
+            tensor = tensor.permute(0, 2, 1, 3).contiguous()
+        else:
+            raise ValueError(f"Input tensor rank should be one of [4, 5], but is: {len(tensor.shape)}")
+        new_shape = tensor.size()[:-2] + (num_attention_heads * attn_head_size,)
+        return tensor.view(new_shape)
+
+    def _attn(
+        self,
+        query,
+        key,
+        value,
+        attention_mask=None,
+        head_mask=None,
+    ):
+
+        # compute causal mask from causal mask buffer
+        query_length, key_length = query.size(-2), key.size(-2)
+        causal_mask = self.bias[:, :, key_length - query_length : key_length, :key_length].to(torch.bool)
+
+        attn_weights = torch.matmul(query, key.transpose(-1, -2))
+
+        mask_value = torch.finfo(attn_weights.dtype).min
+        # Need to be a tensor, otherwise we get error: `RuntimeError: expected scalar type float but found double`.
+        # Need to be on the same device, otherwise `RuntimeError: ..., x and y to be on the same device`
+        mask_value = torch.tensor(mask_value, dtype=attn_weights.dtype).to(attn_weights.device)
+        attn_weights = torch.where(causal_mask, attn_weights, mask_value)
+
+        attn_weights = attn_weights / self.scale_attn
+
+        if attention_mask is not None:
+            # Apply the attention mask
+            attn_weights = attn_weights + attention_mask
+
+        attn_weights = nn.functional.softmax(attn_weights, dim=-1)
+        attn_weights = attn_weights.to(value.dtype)
+        attn_weights = self.attn_dropout(attn_weights)
+
+        # Mask heads if we want to
+        if head_mask is not None:
+            attn_weights = attn_weights * head_mask
+
+        attn_output = torch.matmul(attn_weights, value)
+
+        return attn_output, attn_weights
+
+    def forward(
+        self,
+        hidden_states: Optional[torch.FloatTensor],
+        encoder_hidden_states: Optional[torch.FloatTensor] = None,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        layer_past: Optional[Tuple[torch.Tensor]] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = False,
+        output_attentions: Optional[bool] = False,
+    ) -> Union[
+        Tuple[torch.Tensor, Tuple[torch.Tensor]],
+        Optional[Tuple[torch.Tensor, Tuple[torch.Tensor], Tuple[torch.Tensor, ...]]],
+    ]:
+
+        query = self.q_proj(hidden_states)
+        # if we are doing cross attention
+        if encoder_hidden_states is not None:
+            key = self.k_proj(encoder_hidden_states)
+            value = self.v_proj(encoder_hidden_states)
+        else:
+            key = self.k_proj(hidden_states)
+            value = self.v_proj(hidden_states)
+
+        query = self._split_heads(query, self.num_attention_heads, self.head_dim, True)
+        key = self._split_heads(key, self.num_attention_heads, self.head_dim, True)
+        value = self._split_heads(value, self.num_attention_heads, self.head_dim, False)
+
+        seq_len = key.shape[1]
+        offset = 0
+
+        if layer_past is not None:
+            offset = layer_past[0].shape[-2]
+            seq_len += offset
+
+        if self.rotary_dim is not None:
+            k_rot = key[:, :, :, : self.rotary_dim]
+            k_pass = key[:, :, :, self.rotary_dim :]
+
+            q_rot = query[:, :, :, : self.rotary_dim]
+            q_pass = query[:, :, :, self.rotary_dim :]
+
+            sincos = fixed_pos_embedding(k_rot, 1, seq_len=seq_len)
+            k_rot = apply_rotary_pos_emb(k_rot, sincos, offset=offset)
+            q_rot = apply_rotary_pos_emb(q_rot, sincos, offset=offset)
+
+            key = torch.cat([k_rot, k_pass], dim=-1)
+            query = torch.cat([q_rot, q_pass], dim=-1)
+        else:
+            sincos = fixed_pos_embedding(key, 1, seq_len=seq_len)
+            key = apply_rotary_pos_emb(key, sincos, offset=offset)
+            query = apply_rotary_pos_emb(query, sincos, offset=offset)
+
+        key = key.permute(0, 2, 1, 3)
+        query = query.permute(0, 2, 1, 3)
+
+        if layer_past is not None:
+            past_key = layer_past[0]
+            past_value = layer_past[1]
+            key = torch.cat((past_key, key), dim=-2)
+            value = torch.cat((past_value, value), dim=-2)
+
+        if use_cache is True:
+            present = (key, value)
+        else:
+            present = None
+
+        # compute self-attention: V x Softmax(QK^T)
+        attn_output, attn_weights = self._attn(query, key, value, attention_mask, head_mask)
+
+        attn_output = self._merge_heads(attn_output, self.num_attention_heads, self.head_dim)
+        attn_output = self.out_proj(attn_output)
+        attn_output = self.resid_dropout(attn_output)
+
+        outputs = (attn_output, present)
+        if output_attentions:
+            outputs += (attn_weights,)
+
+        return outputs  # a, present, (attentions)
+
+        
+class GPTJRCrossAttention(GPTJRAttention):
+    def __init__(self, config):
+        super().__init__(config)
+
+        max_positions = config.max_position_embeddings
+        self.register_buffer(
+            "bias",
+            torch.tril(torch.ones((max_positions, max_positions), dtype=torch.uint8)).view(
+                1, 1, max_positions, max_positions
+            ),
+        )
+        self.register_buffer("masked_bias", torch.tensor(-1e9))
+
+        self.attn_dropout = nn.Dropout(config.attn_pdrop)
+        self.resid_dropout = nn.Dropout(config.resid_pdrop)
+
+        self.embed_dim = config.hidden_size
+        self.num_attention_heads = config.num_attention_heads
+        self.head_dim = self.embed_dim // self.num_attention_heads
+        if self.head_dim * self.num_attention_heads != self.embed_dim:
+            raise ValueError(
+                f"embed_dim must be divisible by num_attention_heads (got `embed_dim`: {self.embed_dim} and"
+                f" `num_attention_heads`: {self.num_attention_heads})."
+            )
+        self.scale_attn = torch.sqrt(torch.tensor(self.head_dim, dtype=torch.float32)).to(torch.get_default_dtype())
+
+        self.k_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+        self.v_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+        self.q_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+        self.out_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+
+        
+    def _split_knn_attn_heads(self, tensor, num_attention_heads, attn_head_size):
+        new_shape = tensor.size()[:-1] + (num_attention_heads, attn_head_size)
+        tensor = tensor.view(new_shape)
+
+        return tensor.permute(0, 1, 3, 2)
+
+        
+    def _merge_heads(self, tensor, num_attention_heads, attn_head_size):
+        """
+        Merges attn_head_size dim and num_attn_heads dim into hidden dim
+        """
+        # tensor -> (bs, seq_len, num_attention_heads, head_dim)
+        tensor = tensor.permute(0, 2, 1, 3).contiguous()
+        new_shape = tensor.size()[:-2] + (num_attention_heads * attn_head_size,)
+        return tensor.view(new_shape)
+
+    def _attn(
+        self,
+        query,
+        key,
+        value,
+        attention_mask=None,
+        head_mask=None,
+    ):
+
+        # query -> (bs, num_attention_heads, seq_len, head_dim)
+        # key -> (bs, num_attention_heads, head_dim, neighbors)
+        # attn_weights -> (bs, num_attention_heads, seq_len, neighbors)
+        attn_weights = torch.matmul(query, key)
+
+        attn_weights = nn.functional.softmax(attn_weights, dim=-1)
+        attn_weights = attn_weights.to(value.dtype)
+        attn_weights = self.attn_dropout(attn_weights)
+
+        # Mask heads if we want to
+        if head_mask is not None:
+            attn_weights = attn_weights * head_mask
+
+        # value -> (bs, num_attention_heads, seq_len, head_dim)
+        # attn_weights -> (bs, num_attention_heads, seq_len, neighbors)
+        # attn_output -> (bs, num_attention_heads, seq_len, head_dim)
+        attn_output = torch.matmul(attn_weights, value)
+
+        return attn_output, attn_weights
+
+            
+    def forward(
+        self,
+        hidden_states: Optional[torch.FloatTensor],
+        encoder_hidden_states: Optional[torch.FloatTensor],
+        attention_mask: Optional[torch.FloatTensor] = None,
+        layer_past: Optional[Tuple[torch.Tensor]] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = False,
+        output_attentions: Optional[bool] = False,
+    ) -> Union[
+        Tuple[torch.Tensor, Tuple[torch.Tensor]],
+        Optional[Tuple[torch.Tensor, Tuple[torch.Tensor], Tuple[torch.Tensor, ...]]],
+    ]:
+
+        query = self.q_proj(hidden_states)
+        # if we are doing cross attention
+        key = self.k_proj(encoder_hidden_states)
+        value = self.v_proj(encoder_hidden_states)
+        # (bs, seq_len, dim) -> (bs, num_attention_heads, seq_len, head_dim)
+        query = self._split_heads(query, self.num_attention_heads, self.head_dim, False)
+        # (bs, dim) -> (bs, num_attention_heads,  head_dim)
+        key = self._split_knn_attn_heads(key, self.num_attention_heads, self.head_dim)
+        value = self._split_knn_attn_heads(value, self.num_attention_heads, self.head_dim)
+
+
+        value = value.permute(0, 3, 1, 2)
+        key = key.permute(0, 3, 2, 1)
+
+        if layer_past is not None:
+            past_key = layer_past[0]
+            past_value = layer_past[1]
+            key = torch.cat((past_key, key), dim=-2)
+            value = torch.cat((past_value, value), dim=-2)
+
+        if use_cache is True:
+            present = (key, value)
+        else:
+            present = None
+
+        # compute self-attention: V x Softmax(QK^T)
+        attn_output, attn_weights = self._attn(query, key, value, attention_mask, head_mask)
+
+        attn_output = self._merge_heads(attn_output, self.num_attention_heads, self.head_dim)
+        attn_output = self.out_proj(attn_output)
+        attn_output = self.resid_dropout(attn_output)
+
+        outputs = (attn_output, present)
+        if output_attentions:
+            outputs += (attn_weights,)
+
+        return outputs  # a, present, (attentions)
+        
+
+
+class GPTJRMLP(nn.Module):
+    def __init__(self, intermediate_size, config):  # in MLP: intermediate_size= 4 * embed_dim
+        super().__init__()
+        embed_dim = config.n_embd
+
+        self.fc_in = nn.Linear(embed_dim, intermediate_size)
+        self.fc_out = nn.Linear(intermediate_size, embed_dim)
+
+        self.act = ACT2FN[config.activation_function]
+        self.dropout = nn.Dropout(config.resid_pdrop)
+
+    def forward(self, hidden_states: Optional[torch.FloatTensor]) -> torch.FloatTensor:
+        hidden_states = self.fc_in(hidden_states)
+        hidden_states = self.act(hidden_states)
+        hidden_states = self.fc_out(hidden_states)
+        hidden_states = self.dropout(hidden_states)
+        return hidden_states
+
+
+class GPTJRBlock(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        inner_dim = config.n_inner if config.n_inner is not None else 4 * config.n_embd
+        self.ln_1 = nn.LayerNorm(config.n_embd, eps=config.layer_norm_epsilon)
+        self.attn = GPTJRAttention(config)
+        self.mlp = GPTJRMLP(inner_dim, config)
+
+        self.ln_2 = nn.LayerNorm(config.n_embd, eps=config.layer_norm_epsilon)
+        self.cross_attn = GPTJRCrossAttention(config)
+        self.cross_attn_mlp = GPTJRMLP(inner_dim, config)
+        
+        self.total_alpha_steps = config.total_alpha_steps
+        self.initial_alpha = config.initial_alpha
+        self.final_alpha = config.final_alpha
+
+    def forward(
+        self,
+        hidden_states: Optional[torch.FloatTensor],
+        encoder_hidden_states: Optional[torch.FloatTensor],
+        layer_past: Optional[Tuple[torch.Tensor]] = None,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = False,
+        output_attentions: Optional[bool] = False,
+        step: Optional[int] = None,
+    ) -> Union[Tuple[torch.Tensor], Optional[Tuple[torch.Tensor, Tuple[torch.FloatTensor, ...]]]]:
+        # shape (bs, seq_len, hidden_dim)
+        residual = hidden_states
+        hidden_states = self.ln_1(hidden_states)
+        attn_outputs = self.attn(
+            hidden_states,
+            layer_past=layer_past,
+            attention_mask=attention_mask,
+            head_mask=head_mask,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+        )
+        attn_output = attn_outputs[0]  # output_attn: a, present, (attentions)
+        outputs = attn_outputs[1:]
+
+        feed_forward_hidden_states = self.mlp(hidden_states)
+        self_attention_residual = attn_output + feed_forward_hidden_states + residual
+
+        # encoder_hidden_states -> (bs, knn, encoder_dim)
+        if encoder_hidden_states.dtype != hidden_states.dtype:
+            encoder_hidden_states = encoder_hidden_states.to(hidden_states.dtype)
+        encoder_normed = self.ln_2(encoder_hidden_states)
+
+        # cross_attn_outputs -> (bs, seq_len, dim)
+        cross_attn_output = self.cross_attn(
+            hidden_states,
+            encoder_hidden_states=encoder_normed,
+            attention_mask=attention_mask,
+            head_mask=head_mask,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+        )
+
+        # gpt-j has parallel ff + attn, can do ff on encoder_normed too I guess?
+        cross_attn_ff = self.cross_attn_mlp(
+            cross_attn_output[0] 
+        )
+
+        if step is not None:
+            alpha = self._update_alpha(step)
+        else:
+            alpha = 0.5 
+
+        hidden_states = (1 - alpha) * cross_attn_ff +  alpha * self_attention_residual  
+
+        if use_cache:
+            outputs = (hidden_states,) + outputs
+        else:
+            outputs = (hidden_states,) + outputs[1:]
+
+        return outputs  # hidden_states, present, (attentions)
+        
+    def _update_alpha(self, current_step):
+        """
+        Computes the learning rate for the current step using a cosine decay schedule.
+
+        Args:
+            initial_lr (float): The initial learning rate.
+            final_lr (float): The final learning rate.
+            total_steps (int): The total number of steps in the schedule.
+            current_step (int): The current step.
+
+        Returns:
+            float: The learning rate for the current step.
+        """
+        initial_alpha = 1
+        final_alpha = .5
+        if current_step >= self.total_alpha_steps:
+            return final_alpha
+
+        # Compute the cosine decay factor
+        cosine_decay = 0.5 * (1 + math.cos(math.pi * current_step / self.total_alpha_steps))
+
+        # Compute the current learning rate
+        alpha = final_alpha + (initial_alpha - final_alpha) * cosine_decay
+
+        return alpha
+
+
+class GPTJRPreTrainedModel(PreTrainedModel):
+    """
+    An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained
+    models.
+    """
+
+    config_class = GPTJRConfig
+    base_model_prefix = "transformer"
+    is_parallelizable = True
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["GPTJRBlock"]
+
+    def __init__(self, *inputs, **kwargs):
+        super().__init__(*inputs, **kwargs)
+
+    def _init_weights(self, module):
+        """Initialize the weights."""
+        if isinstance(module, (nn.Linear,)):
+            # Slightly different from Mesh Transformer JAX which uses truncated_normal for initialization
+            # cf https://github.com/pytorch/pytorch/pull/5617
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.Embedding):
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.padding_idx is not None:
+                module.weight.data[module.padding_idx].zero_()
+        elif isinstance(module, nn.LayerNorm):
+            module.bias.data.zero_()
+            module.weight.data.fill_(1.0)
+
+    def _set_gradient_checkpointing(self, module, value=False):
+        if isinstance(module, GPTJRModel):
+            module.gradient_checkpointing = value
+
+
+class GPTJRModel(GPTJRPreTrainedModel):
+    def __init__(self, config):
+        super().__init__(config)
+
+        self.embed_dim = config.n_embd
+        self.vocab_size = config.vocab_size
+        self.wte = nn.Embedding(config.vocab_size, self.embed_dim)
+        self.drop = nn.Dropout(config.embd_pdrop)
+        self.h = nn.ModuleList([GPTJRBlock(config) for _ in range(config.n_layer)])
+        self.ln_f = nn.LayerNorm(self.embed_dim, eps=config.layer_norm_epsilon)
+
+        # Model parallel
+        self.model_parallel = False
+        self.device_map = None
+        self.gradient_checkpointing = False
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def parallelize(self, device_map=None):
+        # Check validity of device_map
+        self.device_map = (
+            get_device_map(len(self.h), range(torch.cuda.device_count())) if device_map is None else device_map
+        )
+        assert_device_map(self.device_map, len(self.h))
+        self.model_parallel = True
+        self.first_device = "cpu" if "cpu" in self.device_map.keys() else "cuda:" + str(min(self.device_map.keys()))
+        self.last_device = "cuda:" + str(max(self.device_map.keys()))
+        self.wte = self.wte.to(self.first_device)
+        # Load onto devices
+        for k, v in self.device_map.items():
+            for block in v:
+                cuda_device = "cuda:" + str(k)
+                self.h[block] = self.h[block].to(cuda_device)
+        # ln_f to last
+        self.ln_f = self.ln_f.to(self.last_device)
+
+    def deparallelize(self):
+        self.model_parallel = False
+        self.device_map = None
+        self.first_device = "cpu"
+        self.last_device = "cpu"
+        self.wte = self.wte.to("cpu")
+        for index in range(len(self.h)):
+            self.h[index] = self.h[index].to("cpu")
+        self.ln_f = self.ln_f.to("cpu")
+        torch.cuda.empty_cache()
+
+    def get_input_embeddings(self):
+        return self.wte
+
+    def set_input_embeddings(self, new_embeddings):
+        self.wte = new_embeddings
+
+    def forward(
+        self,
+        input_ids: Optional[torch.LongTensor] = None,
+        encoder_hidden_states: Optional[torch.FloatTensor] = None,
+        past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        token_type_ids: Optional[torch.LongTensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        step: Optional[int] = None,
+    ) -> Union[Tuple, BaseModelOutputWithPast]:
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if input_ids is not None and inputs_embeds is not None:
+            raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
+        elif input_ids is not None:
+            input_shape = input_ids.size()
+            input_ids = input_ids.view(-1, input_shape[-1])
+            batch_size = input_ids.shape[0]
+        elif inputs_embeds is not None:
+            input_shape = inputs_embeds.size()[:-1]
+            batch_size = inputs_embeds.shape[0]
+        else:
+            raise ValueError("You have to specify either input_ids or inputs_embeds")
+
+        device = input_ids.device if input_ids is not None else inputs_embeds.device
+
+        if token_type_ids is not None:
+            token_type_ids = token_type_ids.view(-1, input_shape[-1])
+
+        if position_ids is not None:
+            position_ids = position_ids.view(-1, input_shape[-1])
+
+        if past_key_values is None:
+            past_length = 0
+            past_key_values = tuple([None] * len(self.h))
+        else:
+            past_length = past_key_values[0][0].size(-2)
+
+        if position_ids is None:
+            position_ids = torch.arange(past_length, input_shape[-1] + past_length, dtype=torch.long, device=device)
+            position_ids = position_ids.unsqueeze(0).view(-1, input_shape[-1])
+
+        # Attention mask.
+        if attention_mask is not None:
+            if batch_size <= 0:
+                raise ValueError("batch_size has to be defined and > 0")
+            attention_mask = attention_mask.view(batch_size, -1)
+            # We create a 3D attention mask from a 2D tensor mask.
+            # Sizes are [batch_size, 1, 1, to_seq_length]
+            # So we can broadcast to [batch_size, num_heads, from_seq_length, to_seq_length]
+            # this attention mask is more simple than the triangular masking of causal attention
+            # used in OpenAI GPT, we just need to prepare the broadcast dimension here.
+            attention_mask = attention_mask[:, None, None, :]
+
+            # Since attention_mask is 1.0 for positions we want to attend and 0.0 for
+            # masked positions, this operation will create a tensor which is 0.0 for
+            # positions we want to attend and the dtype's smallest value for masked positions.
+            # Since we are adding it to the raw scores before the softmax, this is
+            # effectively the same as removing these entirely.
+            attention_mask = attention_mask.to(dtype=self.dtype)  # fp16 compatibility
+            attention_mask = (1.0 - attention_mask) * torch.finfo(self.dtype).min
+
+        # Prepare head mask if needed
+        # 1.0 in head_mask indicate we keep the head
+        # attention_probs has shape bsz x num_attention_heads x N x N
+        # head_mask has shape n_layer x batch x num_attention_heads x N x N
+        head_mask = self.get_head_mask(head_mask, self.config.n_layer)
+
+        if inputs_embeds is None:
+            inputs_embeds = self.wte(input_ids)
+
+        hidden_states = inputs_embeds
+
+        if token_type_ids is not None:
+            token_type_embeds = self.wte(token_type_ids)
+            hidden_states = hidden_states + token_type_embeds
+
+        hidden_states = self.drop(hidden_states)
+
+        output_shape = input_shape + (hidden_states.size(-1),)
+
+        presents = () if use_cache else None
+        all_self_attentions = () if output_attentions else None
+        all_hidden_states = () if output_hidden_states else None
+        for i, (block, layer_past) in enumerate(zip(self.h, past_key_values)):
+
+            # Model parallel
+            if self.model_parallel:
+                torch.cuda.set_device(hidden_states.device)
+                # Ensure layer_past is on same device as hidden_states (might not be correct)
+                if layer_past is not None:
+                    layer_past = tuple(past_state.to(hidden_states.device) for past_state in layer_past)
+                # Ensure that attention_mask is always on the same device as hidden_states
+                if attention_mask is not None:
+                    attention_mask = attention_mask.to(hidden_states.device)
+                if isinstance(head_mask, torch.Tensor):
+                    head_mask = head_mask.to(hidden_states.device)
+            if output_hidden_states:
+                all_hidden_states = all_hidden_states + (hidden_states,)
+
+            if self.gradient_checkpointing and self.training:
+
+                if use_cache:
+                    logger.warning(
+                        "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..."
+                    )
+                    use_cache = False
+
+                def create_custom_forward(module):
+                    def custom_forward(*inputs):
+                        # None for past_key_value
+                        return module(*inputs, use_cache, output_attentions, step)
+
+                    return custom_forward
+
+                outputs = torch.utils.checkpoint.checkpoint(
+                    create_custom_forward(block),
+                    hidden_states,
+                    encoder_hidden_states,
+                    None,
+                    attention_mask,
+                    head_mask[i],
+                )
+            else:
+                outputs = block(
+                    hidden_states,
+                    encoder_hidden_states,
+                    layer_past=layer_past,
+                    attention_mask=attention_mask,
+                    head_mask=head_mask[i],
+                    use_cache=use_cache,
+                    output_attentions=output_attentions,
+                    step=step
+                )
+
+            hidden_states = outputs[0]
+            if use_cache is True:
+                presents = presents + (outputs[1],)
+
+            if output_attentions:
+                all_self_attentions = all_self_attentions + (outputs[2 if use_cache else 1],)
+
+            # Model Parallel: If it's the last layer for that device, put things on the next device
+            if self.model_parallel:
+                for k, v in self.device_map.items():
+                    if i == v[-1] and "cuda:" + str(k) != self.last_device:
+                        hidden_states = hidden_states.to("cuda:" + str(k + 1))
+
+        hidden_states = self.ln_f(hidden_states)
+
+        hidden_states = hidden_states.view(output_shape)
+        # Add last hidden state
+        if output_hidden_states:
+            all_hidden_states = all_hidden_states + (hidden_states,)
+
+        if not return_dict:
+            return tuple(v for v in [hidden_states, presents, all_hidden_states, all_self_attentions] if v is not None)
+
+        return BaseModelOutputWithPast(
+            last_hidden_state=hidden_states,
+            past_key_values=presents,
+            hidden_states=all_hidden_states,
+            attentions=all_self_attentions,
+        )
+
+
+class GPTJRForCausalLM(GPTJRPreTrainedModel):
+    _keys_to_ignore_on_load_missing = [r"h\.\d+\.attn\.masked_bias", r"h\.\d+\.attn\.bias"]
+
+    def __init__(self, config):
+        super().__init__(config)
+        self.transformer = GPTJRModel(config)
+        self.lm_head = nn.Linear(config.n_embd, config.vocab_size)
+
+        self.hidden_size = config.hidden_size
+        self.encoder_dim = config.encoder_dim
+
+        if self.hidden_size != self.encoder_dim:
+            self.enc_dec_proj = nn.Sequential(nn.Linear(config.encoder_dim, config.n_embd * 4),
+                                              nn.Linear(config.n_embd * 4, config.n_embd))
+
+        # Model parallel
+        self.model_parallel = False
+        self.device_map = None
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def parallelize(self, device_map=None):
+        self.device_map = (
+            get_device_map(len(self.transformer.h), range(torch.cuda.device_count()))
+            if device_map is None
+            else device_map
+        )
+        assert_device_map(self.device_map, len(self.transformer.h))
+        self.transformer.parallelize(self.device_map)
+        self.lm_head = self.lm_head.to(self.transformer.first_device)
+        self.model_parallel = True
+
+    def deparallelize(self):
+        self.transformer.deparallelize()
+        self.transformer = self.transformer.to("cpu")
+        self.lm_head = self.lm_head.to("cpu")
+        self.model_parallel = False
+        torch.cuda.empty_cache()
+
+    def get_output_embeddings(self):
+        return self.lm_head
+
+    def set_output_embeddings(self, new_embeddings):
+        self.lm_head = new_embeddings
+
+    def prepare_inputs_for_generation(self, input_ids, past_key_values=None, **kwargs):
+        token_type_ids = kwargs.get("token_type_ids", None)
+        # only last token for inputs_ids if past is defined in kwargs
+        if past_key_values:
+            input_ids = input_ids[:, -1].unsqueeze(-1)
+            if token_type_ids is not None:
+                token_type_ids = token_type_ids[:, -1].unsqueeze(-1)
+
+        attention_mask = kwargs.get("attention_mask", None)
+        position_ids = kwargs.get("position_ids", None)
+
+        if attention_mask is not None and position_ids is None:
+            # create position_ids on the fly for batch generation
+            position_ids = attention_mask.long().cumsum(-1) - 1
+            position_ids.masked_fill_(attention_mask == 0, 1)
+            if past_key_values:
+                position_ids = position_ids[:, -1].unsqueeze(-1)
+        else:
+            position_ids = None
+        return {
+            "input_ids": input_ids,
+            "past_key_values": past_key_values,
+            "use_cache": kwargs.get("use_cache"),
+            "position_ids": position_ids,
+            "attention_mask": attention_mask,
+            "token_type_ids": token_type_ids,
+        }
+
+    def forward(
+        self,
+        input_ids: torch.LongTensor,
+        encoder_hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None,
+        token_type_ids: Optional[torch.LongTensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        step: Optional[int] = None
+    ) -> Union[Tuple, CausalLMOutputWithPast]:
+        r"""
+        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Labels for language modeling. Note that the labels **are shifted** inside the model, i.e. you can set
+            `labels = input_ids` Indices are selected in `[-100, 0, ..., config.vocab_size]` All labels set to `-100`
+            are ignored (masked), the loss is only computed for labels in `[0, ..., config.vocab_size]`
+        """
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if self.hidden_size != self.encoder_dim:
+            encoder_hidden_states = encoder_hidden_states.to(self.enc_dec_proj[0].weight.dtype)
+            encoder_hidden_states = self.enc_dec_proj(encoder_hidden_states)
+
+        transformer_outputs = self.transformer(
+            input_ids,
+            encoder_hidden_states=encoder_hidden_states,
+            past_key_values=past_key_values,
+            attention_mask=attention_mask,
+            token_type_ids=token_type_ids,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            step=step,
+        )
+        hidden_states = transformer_outputs[0]
+
+        # Set device for model parallelism
+        if self.model_parallel:
+            torch.cuda.set_device(self.transformer.first_device)
+            hidden_states = hidden_states.to(self.lm_head.weight.device)
+
+        # make sure sampling in fp16 works correctly and
+        # compute loss in fp32 to match with mesh-tf version
+        # https://github.com/EleutherAI/gpt-neo/blob/89ce74164da2fb16179106f54e2269b5da8db333/models/gpt2/gpt2.py#L179
+        # TODO: do we need to do conversion to fp32 if training in bf16?
+        lm_logits = self.lm_head(hidden_states).to(torch.float32)
+
+        loss = None
+        if labels is not None:
+            # Shift so that tokens < n predict n
+            shift_logits = lm_logits[..., :-1, :].contiguous()
+            shift_labels = labels[..., 1:].contiguous()
+            # Flatten the tokens
+            loss_fct = CrossEntropyLoss()
+            loss = loss_fct(shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1))
+
+            loss = loss.to(hidden_states.dtype)
+
+        if not return_dict:
+            output = (lm_logits,) + transformer_outputs[1:]
+            return ((loss,) + output) if loss is not None else output
+
+        return CausalLMOutputWithPast(
+            loss=loss,
+            logits=lm_logits,
+            past_key_values=transformer_outputs.past_key_values,
+            hidden_states=transformer_outputs.hidden_states,
+            attentions=transformer_outputs.attentions,
+        )
+
+    @staticmethod
+    def _reorder_cache(past: Tuple[Tuple[torch.Tensor]], beam_idx: torch.Tensor) -> Tuple[Tuple[torch.Tensor]]:
+        """
+        This function is used to re-order the `past_key_values` cache if [`~PretrainedModel.beam_search`] or
+        [`~PretrainedModel.beam_sample`] is called. This is required to match `past_key_values` with the correct
+        beam_idx at every generation step.
+        """
+        return tuple(
+            tuple(past_state.index_select(0, beam_idx.to(past_state.device)) for past_state in layer_past)
+            for layer_past in past
+        )
+

gpt4all/models/gpt_jr/test_gpt_jr.py

@@ -0,0 +1,50 @@
+import torch
+from gpt4all.models import GPTJRForCausalLM, GPTJRConfig
+from transformers import AutoTokenizer, AutoModel
+
+# seed torch 
+
+torch.manual_seed(0)
+
+config = GPTJRConfig(encoder_dim=384, n_layer=4)
+print("loaded config")
+
+print("loading model")
+model = GPTJRForCausalLM(config)
+print("loaded model")
+
+
+tokenizer =  AutoTokenizer.from_pretrained("EleutherAI/gpt-j-6b")
+tokenizer.pad_token = tokenizer.eos_token
+
+encoder_tokenizer = AutoTokenizer.from_pretrained('sentence-transformers/all-MiniLM-L6-v2')
+encoder = AutoModel.from_pretrained('sentence-transformers/all-MiniLM-L6-v2')
+
+
+def mean_pooling(model_output, attention_mask):
+    token_embeddings = model_output[0] #First element of model_output contains all token embeddings
+    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
+    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
+
+text = "The quick brown fox jumps over the lazy dog."
+print("Encoded knn")
+tokenized = encoder_tokenizer(text, return_tensors="pt")
+
+# bs, seq_len, dim
+encodings = mean_pooling(encoder(**tokenized), tokenized["attention_mask"])
+
+# make 2 neighbors
+# (bs, knn, encoding_dim)
+encoder_hidden_states = torch.stack([encodings, encodings]).squeeze().unsqueeze(0)
+
+inputs = "What did the fox do?"
+
+print("Encoded inputs")
+tokenized_input = tokenizer([inputs], padding="max_length", truncation=True, return_tensors="pt")
+
+print("Running model")
+outputs = model(**tokenized_input, encoder_hidden_states=encoder_hidden_states)
+
+print(outputs)
+print(outputs[0].shape)
+

gpt4all/models/lethe/__init__.py

@@ -0,0 +1,2 @@
+from .configuration_lethe import LetheConfig
+from .modeling_lethe import LetheForCausalLM

gpt4all/models/lethe/configuration_lethe.py

@@ -0,0 +1,138 @@
+# coding=utf-8
+# Copyright 2022 EleutherAI and The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+""" GPTNeoX model configuration"""
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+GPT_NEOX_PRETRAINED_CONFIG_ARCHIVE_MAP = {
+    "EleutherAI/gpt-neox-20b": "https://huggingface.co/EleutherAI/gpt-neox-20b/resolve/main/config.json",
+    # See all GPTNeoX models at https://huggingface.co/models?filter=gpt_neox
+}
+
+
+class LetheConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`GPTNeoXModel`]. It is used to instantiate an
+    GPTNeoX model according to the specified arguments, defining the model architecture. Instantiating a configuration
+    with the defaults will yield a similar configuration to that of the GPTNeoX
+    [EleutherAI/gpt-neox-20b](https://huggingface.co/EleutherAI/gpt-neox-20b) architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+
+    Args:
+        vocab_size (`int`, *optional*, defaults to 50432):
+            Vocabulary size of the GPTNeoX model. Defines the number of different tokens that can be represented by the
+            `inputs_ids` passed when calling [`GPTNeoXModel`].
+        hidden_size (`int`, *optional*, defaults to 6144):
+            Dimension of the encoder layers and the pooler layer.
+        num_hidden_layers (`int`, *optional*, defaults to 44):
+            Number of hidden layers in the Transformer encoder.
+        num_attention_heads (`int`, *optional*, defaults to 64):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        intermediate_size (`int`, *optional*, defaults to 24576):
+            Dimension of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder.
+        hidden_act (`str` or `function`, *optional*, defaults to `"gelu"`):
+            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
+            `"relu"`, `"selu"` and `"gelu_new"` are supported.
+        rotary_pct (`float`, *optional*, defaults to 0.25):
+            percentage of hidden dimensions to allocate to rotary embeddings
+        rotary_emb_base (`int`, *optional*, defaults to 10000)
+            base for computing rotary embeddings frequency
+        classifier_dropout (`float`, *optional*, defaults to 0.1):
+            Argument used when doing token classification, used in the model [`GPTNeoXForTokenClassification`].
+
+            The dropout ratio for the hidden layer.
+        max_position_embeddings (`int`, *optional*, defaults to 2048):
+            The maximum sequence length that this model might ever be used with. Typically set this to something large
+            just in case (e.g., 512 or 1024 or 2048).
+        initializer_range (`float`, *optional*, defaults to 1e-5):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        layer_norm_eps (`float`, *optional*, defaults to 1e-12):
+            The epsilon used by the layer normalization layers.
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should return the last key/values attentions (not used by all models). Only
+            relevant if `config.is_decoder=True`.
+        use_parallel_residual (`bool`, *optional*, defaults to `True`):
+            Whether to use a "parallel" formulation in each Transformer layer, which can provide a slight training
+            speedup at large scales (e.g. 20B).
+        Example:
+
+    ```python
+    >>> from transformers import GPTNeoXConfig, GPTNeoXModel
+
+    >>> # Initializing a GPTNeoX gpt-neox-20b style configuration
+    >>> configuration = GPTNeoXConfig()
+
+    >>> # Initializing a model (with random weights) from the gpt-neox-20b style configuration
+    >>> model = GPTNeoXModel(configuration)  # doctest: +SKIP
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config  # doctest: +SKIP
+    ```"""
+    model_type = "gpt_neox"
+
+    def __init__(
+        self,
+        vocab_size=50432,
+        hidden_size=6144,
+        num_hidden_layers=44,
+        num_attention_heads=64,
+        intermediate_size=24576,
+        hidden_act="gelu",
+        rotary_pct=0.25,
+        rotary_emb_base=10000,
+        classifier_dropout=0.1,
+        max_position_embeddings=2048,
+        initializer_range=0.02,
+        layer_norm_eps=1e-5,
+        use_cache=True,
+        bos_token_id=0,
+        eos_token_id=2,
+        tie_word_embeddings=False,
+        use_parallel_residual=True,
+        memory_attn_layer=9,
+        num_neighbors_to_retrieve=32,
+        num_neighbors_stored=128,
+        attn_scale_init=20.0,
+        **kwargs,
+    ):
+        super().__init__(bos_token_id=bos_token_id, eos_token_id=eos_token_id, **kwargs)
+        self.vocab_size = vocab_size
+        self.max_position_embeddings = max_position_embeddings
+        self.hidden_size = hidden_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.intermediate_size = intermediate_size
+        self.hidden_act = hidden_act
+        self.rotary_pct = rotary_pct
+        self.rotary_emb_base = rotary_emb_base
+        self.classifier_dropout = classifier_dropout
+        self.initializer_range = initializer_range
+        self.layer_norm_eps = layer_norm_eps
+        self.use_cache = use_cache
+        self.tie_word_embeddings = tie_word_embeddings
+        self.use_parallel_residual = use_parallel_residual
+
+        # index of cross attention layer to add
+        self.memory_attn_layer = memory_attn_layer 
+        self.num_neighbors_to_retrieve = num_neighbors_to_retrieve
+        self.num_neighbors_stored = num_neighbors_stored
+        self.attn_scale_init = attn_scale_init

gpt4all/models/lethe/test_index.py

@@ -0,0 +1,21 @@
+import time
+import numpy as np
+from gpt4all.models.lethe.modeling_lethe import MemoryIndex
+
+
+index = MemoryIndex(256,
+                    575000,
+                    8 
+)
+
+keys = np.random.randn(32, 8, 1024, 256)
+values = np.random.randn(32, 8, 1024, 256)
+start = time.time()
+index.add(keys, values)
+print(f"index.add time: {time.time() - start}")
+
+print(index.key_indices[0].index.ntotal)
+queries = np.random.randn(32, 8, 1024, 256)
+start = time.time()
+index.knn_query(queries, k=32)
+print(f"index.knn_query time: {time.time() - start}")

gpt4all/models/lethe/test_lethe.py

@@ -0,0 +1,84 @@
+import torch
+from gpt4all.models import LetheForCausalLM, LetheConfig
+from gpt4all.models.lethe.modeling_lethe import MemoryIndex
+from transformers import AutoTokenizer, AutoModel
+from datasets import load_from_disk
+from tqdm import tqdm
+
+# seed torch 
+
+torch.manual_seed(0)
+
+config = LetheConfig(num_hidden_layers=15, 
+                     hidden_size=2048, 
+                     intermediate_size=8192, 
+                     num_attention_heads=8,
+                     memory_attn_layer=12,
+                     num_neighbors_stored=6_000_000,
+                     num_neighbors_to_retrieve=32,
+)
+print("loaded config")
+
+print("loading model")
+dimension = config.max_position_embeddings * config.hidden_size
+head_size = config.hidden_size // config.num_attention_heads
+index = MemoryIndex(head_size,
+                    5_000_000,
+                    # 2 since multi-query attention and storing one each for key and value
+                    config.num_attention_heads,
+)
+model = LetheForCausalLM(config, index)
+model.to("cuda:0")
+print("loaded model")
+
+
+tokenizer =  AutoTokenizer.from_pretrained("EleutherAI/pythia-1b")
+tokenizer.pad_token = tokenizer.eos_token
+tokenizer.model_max_length = 2048
+
+dataset = load_from_disk("/home/paperspace/gpt4all/gpt4all/index/squad_supplemented_train")
+
+item = dataset[0]
+
+question = item["question"]
+answer = item["answers"]["text"][0]
+print(f"question: {question}")
+print(f"answer: {answer}")
+
+contexts = item["neighbor_text"]
+
+contexts_encoded = tokenizer(contexts, padding="max_length", truncation=True, return_tensors="pt")
+tokenized_input = tokenizer(question + "\n" + answer, return_tensors="pt", padding="max_length", truncation=True)
+
+inputs = torch.concatenate([contexts_encoded["input_ids"], tokenized_input["input_ids"]], axis=0)
+token_type_ids = torch.tensor([[0] * len(contexts_encoded["input_ids"]) + [1] * len(tokenized_input["input_ids"])]).squeeze()
+
+question_len = len(tokenizer(question + "\n", return_tensors="pt")["input_ids"][0])
+
+labels = inputs.clone()
+labels[:-1] = -100
+labels[-1, :question_len] = -100
+
+
+memory_mask = token_type_ids == 0
+# should be shape (num_memories, sequence_length)
+memories = inputs[memory_mask]
+
+print("Running model")
+model.eval()
+with torch.no_grad():
+    for chunk_start in tqdm(range(0, memories.shape[0], 32)):
+        chunk_end = min(memories.shape[0], chunk_start + 32)
+        mem_chunk = memories[chunk_start:chunk_end].to(model.device)
+        model(input_ids=mem_chunk, labels=None,)
+
+model.train()
+
+qa_inputs = inputs[~memory_mask]
+qa_labels = labels[~memory_mask]
+outputs = model(input_ids=qa_inputs.to(model.device), labels=qa_labels.to(model.device))
+
+print(outputs)
+print(outputs.logits.shape)
+
+index.reset()

gpt4all/models/pythia_retro/__init__.py

@@ -0,0 +1,2 @@
+from .configuration_pythia_retro import PythiaRetroConfig
+from .modeling_pythia_retro import PythiaRetroForCausalLM

gpt4all/models/pythiaseek/__init__.py

@@ -0,0 +1,2 @@
+from .configuration_pythiaseek import PythiaSeekConfig
+from .modeling_pythiaseek import PythiaSeekForCausalLM

gpt4all/models/pythiaseek/configuration_pythiaseek.py

@@ -0,0 +1,142 @@
+# coding=utf-8
+# Copyright 2022 EleutherAI and The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+""" GPTNeoX model configuration"""
+
+from transformers.configuration_utils import PretrainedConfig
+from transformers.utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+GPT_NEOX_PRETRAINED_CONFIG_ARCHIVE_MAP = {
+    "EleutherAI/gpt-neox-20b": "https://huggingface.co/EleutherAI/gpt-neox-20b/resolve/main/config.json",
+    # See all GPTNeoX models at https://huggingface.co/models?filter=gpt_neox
+}
+
+
+class PythiaSeekConfig(PretrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`GPTNeoXModel`]. It is used to instantiate an
+    GPTNeoX model according to the specified arguments, defining the model architecture. Instantiating a configuration
+    with the defaults will yield a similar configuration to that of the GPTNeoX
+    [EleutherAI/gpt-neox-20b](https://huggingface.co/EleutherAI/gpt-neox-20b) architecture.
+
+    Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PretrainedConfig`] for more information.
+
+
+    Args:
+        vocab_size (`int`, *optional*, defaults to 50432):
+            Vocabulary size of the GPTNeoX model. Defines the number of different tokens that can be represented by the
+            `inputs_ids` passed when calling [`GPTNeoXModel`].
+        hidden_size (`int`, *optional*, defaults to 6144):
+            Dimension of the encoder layers and the pooler layer.
+        num_hidden_layers (`int`, *optional*, defaults to 44):
+            Number of hidden layers in the Transformer encoder.
+        num_attention_heads (`int`, *optional*, defaults to 64):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        intermediate_size (`int`, *optional*, defaults to 24576):
+            Dimension of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder.
+        hidden_act (`str` or `function`, *optional*, defaults to `"gelu"`):
+            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
+            `"relu"`, `"selu"` and `"gelu_new"` are supported.
+        rotary_pct (`float`, *optional*, defaults to 0.25):
+            percentage of hidden dimensions to allocate to rotary embeddings
+        rotary_emb_base (`int`, *optional*, defaults to 10000)
+            base for computing rotary embeddings frequency
+        classifier_dropout (`float`, *optional*, defaults to 0.1):
+            Argument used when doing token classification, used in the model [`GPTNeoXForTokenClassification`].
+
+            The dropout ratio for the hidden layer.
+        max_position_embeddings (`int`, *optional*, defaults to 2048):
+            The maximum sequence length that this model might ever be used with. Typically set this to something large
+            just in case (e.g., 512 or 1024 or 2048).
+        initializer_range (`float`, *optional*, defaults to 1e-5):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        layer_norm_eps (`float`, *optional*, defaults to 1e-12):
+            The epsilon used by the layer normalization layers.
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should return the last key/values attentions (not used by all models). Only
+            relevant if `config.is_decoder=True`.
+        use_parallel_residual (`bool`, *optional*, defaults to `True`):
+            Whether to use a "parallel" formulation in each Transformer layer, which can provide a slight training
+            speedup at large scales (e.g. 20B).
+        Example:
+
+    ```python
+    >>> from transformers import GPTNeoXConfig, GPTNeoXModel
+
+    >>> # Initializing a GPTNeoX gpt-neox-20b style configuration
+    >>> configuration = GPTNeoXConfig()
+
+    >>> # Initializing a model (with random weights) from the gpt-neox-20b style configuration
+    >>> model = GPTNeoXModel(configuration)  # doctest: +SKIP
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config  # doctest: +SKIP
+    ```"""
+    model_type = "gpt_neox"
+
+    def __init__(
+        self,
+        vocab_size=50432,
+        hidden_size=6144,
+        num_hidden_layers=44,
+        num_attention_heads=64,
+        intermediate_size=24576,
+        hidden_act="gelu",
+        rotary_pct=0.25,
+        rotary_emb_base=10000,
+        classifier_dropout=0.1,
+        max_position_embeddings=2048,
+        initializer_range=0.02,
+        layer_norm_eps=1e-5,
+        use_cache=True,
+        bos_token_id=0,
+        eos_token_id=2,
+        tie_word_embeddings=False,
+        use_parallel_residual=True,
+        encoder_dim=4096,
+        total_alpha_steps=0,
+        initial_alpha=1,
+        final_alpha=.5,
+        cross_attn_layer=9,
+        learnable_alpha=False,
+        **kwargs,
+    ):
+        super().__init__(bos_token_id=bos_token_id, eos_token_id=eos_token_id, **kwargs)
+        self.vocab_size = vocab_size
+        self.max_position_embeddings = max_position_embeddings
+        self.hidden_size = hidden_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.intermediate_size = intermediate_size
+        self.hidden_act = hidden_act
+        self.rotary_pct = rotary_pct
+        self.rotary_emb_base = rotary_emb_base
+        self.classifier_dropout = classifier_dropout
+        self.initializer_range = initializer_range
+        self.layer_norm_eps = layer_norm_eps
+        self.use_cache = use_cache
+        self.tie_word_embeddings = tie_word_embeddings
+        self.use_parallel_residual = use_parallel_residual
+
+        self.encoder_dim = encoder_dim
+        self.total_alpha_steps = total_alpha_steps
+        self.initial_alpha = initial_alpha
+        self.final_alpha = final_alpha
+        # index of cross attention layer to add
+        self.cross_attn_layer = cross_attn_layer
+        self.learnable_alpha = learnable_alpha

gpt4all/models/pythiaseek/modeling_pythiaseek.py

@@ -0,0 +1,899 @@
+# coding=utf-8
+# Copyright 2022 EleutherAI The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+""" PyTorch PythiaSeek model."""
+
+from typing import Optional, Tuple, Union
+
+import math
+import torch.nn.functional as F
+import torch
+import torch.utils.checkpoint
+from torch import nn
+from torch.nn import CrossEntropyLoss
+
+from transformers import AutoModel
+from transformers.activations import ACT2FN
+from transformers.modeling_outputs import (
+    BaseModelOutputWithPast,
+    CausalLMOutputWithPast,
+)
+from transformers.modeling_utils import PreTrainedModel
+from transformers.utils import logging
+from transformers.utils.model_parallel_utils import assert_device_map, get_device_map
+from gpt4all.models.pythiaseek import PythiaSeekConfig
+
+
+logger = logging.get_logger(__name__)
+
+
+GPT_NEOX_PRETRAINED_MODEL_ARCHIVE_LIST = [
+    "EleutherAI/gpt-neox-20b",
+    # See all PythiaSeek models at https://huggingface.co/models?filter=gpt_neox
+]
+
+
+class PythiaSeekPreTrainedModel(PreTrainedModel):
+    """
+    An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained
+    models.
+    """
+
+    config_class = PythiaSeekConfig
+    base_model_prefix = "gpt_neox"
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["PythiaSeekLayer"]
+
+    def _init_weights(self, module):
+        """Initialize the weights"""
+        if isinstance(module, nn.Linear):
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.Embedding):
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.padding_idx is not None:
+                module.weight.data[module.padding_idx].zero_()
+        elif isinstance(module, nn.LayerNorm):
+            module.bias.data.zero_()
+            module.weight.data.fill_(1.0)
+
+    def _set_gradient_checkpointing(self, module, value=False):
+        if isinstance(module, PythiaSeekModel):
+            module.gradient_checkpointing = value
+
+
+class PythiaSeekAttention(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.num_attention_heads = config.num_attention_heads
+        self.hidden_size = config.hidden_size
+        self.head_size = self.hidden_size // self.num_attention_heads
+        self.rotary_ndims = int(self.head_size * config.rotary_pct)
+        max_positions = config.max_position_embeddings
+        self.register_buffer(
+            "bias",
+            torch.tril(torch.ones((max_positions, max_positions), dtype=torch.bool)).view(
+                1, 1, max_positions, max_positions
+            ),
+        )
+        self.register_buffer("masked_bias", torch.tensor(-1e9))
+        self.rotary_emb = RotaryEmbedding(
+            self.rotary_ndims, config.max_position_embeddings, base=config.rotary_emb_base
+        )
+        self.register_buffer(
+            "norm_factor",
+            torch.sqrt(torch.tensor(self.head_size, dtype=torch.float32)).to(torch.get_default_dtype()),
+            persistent=False,
+        )
+        self.query_key_value = nn.Linear(config.hidden_size, 3 * config.hidden_size)
+        self.dense = nn.Linear(config.hidden_size, config.hidden_size)
+
+    def forward(
+        self,
+        hidden_states: torch.FloatTensor,
+        attention_mask: torch.FloatTensor,
+        position_ids: torch.LongTensor,
+        head_mask: Optional[torch.FloatTensor] = None,
+        layer_past: Optional[Tuple[torch.Tensor]] = None,
+        use_cache: Optional[bool] = False,
+        output_attentions: Optional[bool] = False,
+    ):
+        has_layer_past = layer_past is not None
+
+        # Compute QKV
+        # Attention heads [batch, seq_len, hidden_size]
+        #   --> [batch, seq_len, (np * 3 * head_size)]
+        qkv = self.query_key_value(hidden_states)
+
+        # [batch, seq_len, (num_heads * 3 * head_size)]
+        #   --> [batch, seq_len, num_heads, 3 * head_size]
+        new_qkv_shape = qkv.size()[:-1] + (self.num_attention_heads, 3 * self.head_size)
+        qkv = qkv.view(*new_qkv_shape)
+
+        # [batch, seq_len, num_attention_heads, 3 * head_size] --> 3 [batch, num_attention_heads, seq_len, head_size]
+        query = qkv[..., : self.head_size].permute(0, 2, 1, 3)
+        key = qkv[..., self.head_size : 2 * self.head_size].permute(0, 2, 1, 3)
+        value = qkv[..., 2 * self.head_size :].permute(0, 2, 1, 3)
+
+        # Compute rotary embeddings on rotary_ndims
+        query_rot = query[..., : self.rotary_ndims]
+        query_pass = query[..., self.rotary_ndims :]
+        key_rot = key[..., : self.rotary_ndims]
+        key_pass = key[..., self.rotary_ndims :]
+
+        # Compute token offset for rotary embeddings (when decoding)
+        seq_len = key.shape[-2]
+        if has_layer_past:
+            seq_len += layer_past[0].shape[-2]
+        cos, sin = self.rotary_emb(value, seq_len=seq_len)
+        query, key = apply_rotary_pos_emb(query_rot, key_rot, cos, sin, position_ids)
+        query = torch.cat((query, query_pass), dim=-1)
+        key = torch.cat((key, key_pass), dim=-1)
+
+        # Cache QKV values
+        if has_layer_past:
+            past_key = layer_past[0]
+            past_value = layer_past[1]
+            key = torch.cat((past_key, key), dim=-2)
+            value = torch.cat((past_value, value), dim=-2)
+        present = (key, value) if use_cache else None
+
+        # Compute attention
+        attn_output, attn_weights = self._attn(query, key, value, attention_mask, head_mask)
+
+        # Reshape outputs
+        attn_output = self._merge_heads(attn_output, self.num_attention_heads, self.head_size)
+        attn_output = self.dense(attn_output)
+
+        outputs = (attn_output, present)
+        if output_attentions:
+            outputs += (attn_weights,)
+
+        return outputs
+
+    @classmethod
+    def _split_heads(cls, tensor, num_attention_heads, attn_head_size):
+        """
+        Splits hidden dim into attn_head_size and num_attention_heads
+        """
+        # tensor: [bs, seq_len, hidden_size]
+        new_shape = tensor.size()[:-1] + (num_attention_heads, attn_head_size)
+        # -> [bs, seq_len, num_attention_heads, attn_head_size]
+        tensor = tensor.view(new_shape)
+        # -> [bs, num_attention_heads, seq_len, attn_head_size]
+        tensor = tensor.permute(0, 2, 1, 3)
+        return tensor
+
+    @classmethod
+    def _merge_heads(cls, tensor, num_attention_heads, attn_head_size):
+        """
+        Merges attn_head_size dim and num_attn_heads dim into hidden dim
+        """
+        # tensor [bs, num_attention_heads, seq_len, attn_head_size]
+        tensor = tensor.permute(0, 2, 1, 3).contiguous()
+        # -> [bs, seq_len, num_attention_heads, attn_head_size]
+        tensor = tensor.view(tensor.size(0), tensor.size(1), num_attention_heads * attn_head_size)
+        # -> [bs, seq_len, hidden_size]
+        return tensor
+
+    def _attn(self, query, key, value, attention_mask=None, head_mask=None):
+        # q, k, v: [bs, num_attention_heads, seq_len, attn_head_size]
+        # compute causal mask from causal mask buffer
+        batch_size, num_attention_heads, query_length, attn_head_size = query.size()
+        key_length = key.size(-2)
+
+        causal_mask = self.bias[:, :, key_length - query_length : key_length, :key_length]
+
+        query = query.view(batch_size * num_attention_heads, query_length, attn_head_size)
+        key = key.view(batch_size * num_attention_heads, key_length, attn_head_size)
+        attn_scores = torch.zeros(
+            batch_size * num_attention_heads,
+            query_length,
+            key_length,
+            dtype=query.dtype,
+            device=key.device,
+        )
+        attn_scores = torch.baddbmm(
+            attn_scores,
+            query,
+            key.transpose(1, 2),
+            beta=1.0,
+            alpha=(torch.tensor(1.0, dtype=self.norm_factor.dtype, device=self.norm_factor.device) / self.norm_factor),
+        )
+        attn_scores = attn_scores.view(batch_size, num_attention_heads, query_length, key_length)
+
+        mask_value = torch.finfo(attn_scores.dtype).min
+        # Need to be a tensor, otherwise we get error: `RuntimeError: expected scalar type float but found double`.
+        # Need to be on the same device, otherwise `RuntimeError: ..., x and y to be on the same device`
+        mask_value = torch.tensor(mask_value, dtype=attn_scores.dtype).to(attn_scores.device)
+        attn_scores = torch.where(causal_mask, attn_scores, mask_value)
+
+        if attention_mask is not None:
+            # Apply the attention mask
+            attn_scores = attn_scores + attention_mask
+
+        attn_weights = nn.functional.softmax(attn_scores, dim=-1)
+        attn_weights = attn_weights.to(value.dtype)
+
+        # Mask heads if we want to
+        if head_mask is not None:
+            attn_weights = attn_weights * head_mask
+
+        attn_output = torch.matmul(attn_weights, value)
+        return attn_output, attn_weights
+
+class PythiaSeekCrossAttention(PythiaSeekAttention):
+    def __init__(self, config):
+        super().__init__(config)
+
+        self.embed_dim = config.hidden_size
+        self.num_attention_heads = config.num_attention_heads
+        self.head_dim = self.embed_dim // self.num_attention_heads
+        if self.head_dim * self.num_attention_heads != self.embed_dim:
+            raise ValueError(
+                f"embed_dim must be divisible by num_attention_heads (got `embed_dim`: {self.embed_dim} and"
+                f" `num_attention_heads`: {self.num_attention_heads})."
+            )
+        self.norm_factor = torch.sqrt(torch.tensor(self.head_dim, dtype=torch.float32)).to(torch.get_default_dtype())
+
+        self.k_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+        self.v_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+        self.q_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+        self.out_proj = nn.Linear(self.embed_dim, self.embed_dim, bias=False)
+
+        
+    def _split_knn_attn_heads(self, tensor, num_attention_heads, attn_head_size):
+        new_shape = tensor.size()[:-1] + (num_attention_heads, attn_head_size)
+        tensor = tensor.view(new_shape)
+
+        return tensor.permute(0, 1, 3, 2)
+
+        
+    def _merge_heads(self, tensor, num_attention_heads, attn_head_size):
+        """
+        Merges attn_head_size dim and num_attn_heads dim into hidden dim
+        """
+        # tensor -> (bs, seq_len, num_attention_heads, head_dim)
+        tensor = tensor.permute(0, 2, 1, 3).contiguous()
+        new_shape = tensor.size()[:-2] + (num_attention_heads * attn_head_size,)
+        return tensor.view(new_shape)
+
+    def _attn(
+        self,
+        query,
+        key,
+        value,
+        attention_mask=None,
+        head_mask=None,
+    ):
+
+        # query -> (bs, num_attention_heads, seq_len, head_dim)
+        # key -> (bs, num_attention_heads, head_dim, neighbors)
+        # attn_weights -> (bs, num_attention_heads, seq_len, neighbors)
+        attn_weights = torch.matmul(query, key) / self.norm_factor
+
+        attn_weights = nn.functional.softmax(attn_weights, dim=-1)
+        attn_weights = attn_weights.to(value.dtype)
+
+        # Mask heads if we want to
+        if head_mask is not None:
+            attn_weights = attn_weights * head_mask
+
+        # value -> (bs, num_attention_heads, seq_len, head_dim)
+        # attn_weights -> (bs, num_attention_heads, seq_len, neighbors)
+        # attn_output -> (bs, num_attention_heads, seq_len, head_dim)
+        attn_output = torch.matmul(attn_weights, value)
+
+        return attn_output, attn_weights
+
+            
+    def forward(
+        self,
+        hidden_states: Optional[torch.FloatTensor],
+        encoder_hidden_states: Optional[torch.FloatTensor],
+        attention_mask: Optional[torch.FloatTensor] = None,
+        layer_past: Optional[Tuple[torch.Tensor]] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = False,
+        output_attentions: Optional[bool] = False,
+    ) -> Union[
+        Tuple[torch.Tensor, Tuple[torch.Tensor]],
+        Optional[Tuple[torch.Tensor, Tuple[torch.Tensor], Tuple[torch.Tensor, ...]]],
+    ]:
+
+        query = self.q_proj(hidden_states)
+        # if we are doing cross attention
+        key = self.k_proj(encoder_hidden_states)
+        value = self.v_proj(encoder_hidden_states)
+        # (bs, seq_len, dim) -> (bs, num_attention_heads, seq_len, head_dim)
+        query = self._split_heads(query, self.num_attention_heads, self.head_dim)
+        # (bs, dim) -> (bs, num_attention_heads,  head_dim)
+        key = self._split_knn_attn_heads(key, self.num_attention_heads, self.head_dim)
+        value = self._split_knn_attn_heads(value, self.num_attention_heads, self.head_dim)
+
+
+        value = value.permute(0, 3, 1, 2)
+        key = key.permute(0, 3, 2, 1)
+
+        if layer_past is not None:
+            past_key = layer_past[0]
+            past_value = layer_past[1]
+            key = torch.cat((past_key, key), dim=-2)
+            value = torch.cat((past_value, value), dim=-2)
+
+        if use_cache is True:
+            present = (key, value)
+        else:
+            present = None
+
+        # compute self-attention: V x Softmax(QK^T)
+        attn_output, attn_weights = self._attn(query, key, value, attention_mask, head_mask)
+
+        attn_output = self._merge_heads(attn_output, self.num_attention_heads, self.head_dim)
+        attn_output = self.out_proj(attn_output)
+
+        outputs = (attn_output, present)
+        if output_attentions:
+            outputs += (attn_weights,)
+
+        return outputs  # a, present, (attentions)
+
+
+def attention_mask_func(attention_scores, ltor_mask):
+    attention_scores.masked_fill_(~ltor_mask, torch.finfo(attention_scores.dtype).min)
+    return attention_scores
+
+
+class RotaryEmbedding(torch.nn.Module):
+    def __init__(self, dim, max_position_embeddings, base=10000, device=None):
+        super().__init__()
+        inv_freq = 1.0 / (base ** (torch.arange(0, dim, 2).float().to(device) / dim))
+        self.register_buffer("inv_freq", inv_freq)
+
+        # Build here to make `torch.jit.trace` work.
+        self.max_seq_len_cached = max_position_embeddings
+        t = torch.arange(self.max_seq_len_cached, device=self.inv_freq.device, dtype=self.inv_freq.dtype)
+        freqs = torch.einsum("i,j->ij", t, self.inv_freq)
+        # Different from paper, but it uses a different permutation in order to obtain the same calculation
+        emb = torch.cat((freqs, freqs), dim=-1)
+        self.cos_cached = emb.cos()[None, None, :, :]
+        self.sin_cached = emb.sin()[None, None, :, :]
+
+    def forward(self, x, seq_len=None):
+        # x: [bs, num_attention_heads, seq_len, head_size]
+        # This `if` block is unlikely to be run after we build sin/cos in `__init__`. Keep the logic here just in case.
+        if seq_len > self.max_seq_len_cached:
+            self.max_seq_len_cached = seq_len
+            t = torch.arange(self.max_seq_len_cached, device=x.device, dtype=self.inv_freq.dtype)
+            freqs = torch.einsum("i,j->ij", t, self.inv_freq)
+            # Different from paper, but it uses a different permutation in order to obtain the same calculation
+            emb = torch.cat((freqs, freqs), dim=-1).to(x.device)
+            self.cos_cached = emb.cos()[None, None, :, :]
+            self.sin_cached = emb.sin()[None, None, :, :]
+        return self.cos_cached[:seq_len, ...].to(x.device), self.sin_cached[:seq_len, ...].to(x.device)
+
+
+def rotate_half(x):
+    """Rotates half the hidden dims of the input."""
+    x1 = x[..., : x.shape[-1] // 2]
+    x2 = x[..., x.shape[-1] // 2 :]
+    return torch.cat((-x2, x1), dim=-1)
+
+
+def apply_rotary_pos_emb(q, k, cos, sin, position_ids):
+    gather_indices = position_ids[:, None, :, None]  # [bs, 1, seq_len, 1]
+    gather_indices = gather_indices.repeat(1, cos.shape[1], 1, cos.shape[3])
+    cos = torch.gather(cos.repeat(gather_indices.shape[0], 1, 1, 1), 2, gather_indices)
+    sin = torch.gather(sin.repeat(gather_indices.shape[0], 1, 1, 1), 2, gather_indices)
+    q_embed = (q * cos) + (rotate_half(q) * sin)
+    k_embed = (k * cos) + (rotate_half(k) * sin)
+    return q_embed, k_embed
+
+
+class PythiaSeekMLP(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.dense_h_to_4h = nn.Linear(config.hidden_size, config.intermediate_size)
+        self.dense_4h_to_h = nn.Linear(config.intermediate_size, config.hidden_size)
+        self.act = ACT2FN[config.hidden_act]
+
+    def forward(self, hidden_states):
+        hidden_states = self.dense_h_to_4h(hidden_states)
+        hidden_states = self.act(hidden_states)
+        hidden_states = self.dense_4h_to_h(hidden_states)
+        return hidden_states
+
+
+class PythiaSeekLayer(nn.Module):
+    def __init__(self, config, add_cross_attention=False):
+        super().__init__()
+        self.use_parallel_residual = config.use_parallel_residual
+        self.input_layernorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+        self.post_attention_layernorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+        self.attention = PythiaSeekAttention(config)
+        self.mlp = PythiaSeekMLP(config)
+
+        self.add_cross_attention = add_cross_attention
+
+        if add_cross_attention:
+            self.cross_attn_ln = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+            self.cross_attn = PythiaSeekCrossAttention(config)
+            self.cross_attn_mlp = PythiaSeekMLP(config)
+            
+            self.total_alpha_steps = config.total_alpha_steps
+            self.initial_alpha = config.initial_alpha
+            self.final_alpha = config.final_alpha
+            self.learnable_alpha = config.learnable_alpha
+            if self.learnable_alpha:
+                self.alpha = nn.Parameter(torch.zeros(1), requires_grad=True)
+
+    def forward(
+        self,
+        hidden_states: Optional[torch.FloatTensor],
+        encoder_hidden_states: Optional[torch.FloatTensor],
+        attention_mask: Optional[torch.FloatTensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = False,
+        layer_past: Optional[Tuple[torch.Tensor]] = None,
+        output_attentions: Optional[bool] = False,
+        step: Optional[int] = None,
+    ):
+        ln_hidden_states = self.input_layernorm(hidden_states)
+        attention_layer_outputs = self.attention(
+            ln_hidden_states,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            layer_past=layer_past,
+            head_mask=head_mask,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+        )
+        attn_output = attention_layer_outputs[0]  # output_attn: attn_output, present, (attn_weights)
+        outputs = attention_layer_outputs[1:]
+
+        if self.use_parallel_residual:
+            # pseudocode:
+            # x = x + attn(ln1(x)) + mlp(ln2(x))
+            mlp_output = self.mlp(self.post_attention_layernorm(hidden_states))
+            self_attention_residual = mlp_output + attn_output + hidden_states
+        else:
+            # pseudocode:
+            # x = x + attn(ln1(x))
+            # x = x + mlp(ln2(x))
+            attn_output = attn_output + hidden_states
+            mlp_output = self.mlp(self.post_attention_layernorm(attn_output))
+            self_attention_residual = mlp_output + attn_output
+
+        # encoder_hidden_states -> (bs, knn, encoder_dim)
+        if encoder_hidden_states.dtype != ln_hidden_states.dtype:
+            encoder_hidden_states = encoder_hidden_states.to(ln_hidden_states.dtype)
+
+        if self.add_cross_attention:
+            encoder_normed = self.cross_attn_ln(encoder_hidden_states)
+
+            # cross_attn_outputs -> (bs, seq_len, dim)
+            cross_attn_output = self.cross_attn(
+                ln_hidden_states,
+                encoder_hidden_states=encoder_normed,
+                attention_mask=attention_mask,
+                head_mask=head_mask,
+                use_cache=use_cache,
+                output_attentions=output_attentions,
+            )
+
+            cross_attn_ff = self.cross_attn_mlp(
+                cross_attn_output[0] 
+            )
+
+            if self.learnable_alpha:
+                alpha = F.sigmoid(self.alpha)
+            elif step is not None:
+                alpha = self._update_alpha(step)
+            else:
+                alpha = 0.5 
+
+            hidden_states = (1 - alpha) * cross_attn_ff +  alpha * self_attention_residual  
+        else:
+            hidden_states = self_attention_residual
+
+        if use_cache:
+            outputs = (hidden_states,) + outputs
+        else:
+            outputs = (hidden_states,) + outputs[1:]
+
+        return outputs  # hidden_states, present, (attentions)
+        
+    def _update_alpha(self, current_step):
+        """
+        Computes the learning rate for the current step using a cosine decay schedule.
+
+        Args:
+            initial_lr (float): The initial learning rate.
+            final_lr (float): The final learning rate.
+            total_steps (int): The total number of steps in the schedule.
+            current_step (int): The current step.
+
+        Returns:
+            float: The learning rate for the current step.
+        """
+        initial_alpha = 1
+        final_alpha = .5
+        if current_step >= self.total_alpha_steps:
+            return final_alpha
+
+        # Compute the cosine decay factor
+        cosine_decay = 0.5 * (1 + math.cos(math.pi * current_step / self.total_alpha_steps))
+
+        # Compute the current learning rate
+        alpha = final_alpha + (initial_alpha - final_alpha) * cosine_decay
+
+        return alpha
+
+
+
+class PythiaSeekModel(PythiaSeekPreTrainedModel):
+    def __init__(self, config):
+        super().__init__(config)
+        self.config = config
+
+        self.embed_in = nn.Embedding(config.vocab_size, config.hidden_size)
+
+        if isinstance(config.cross_attn_layer, int):
+            cross_attn_layers = [config.cross_attn_layer]
+        # if we add cross attention to all layers
+        elif config.cross_attn_layer is None:
+            cross_attn_layers = list(range(config.num_hidden_layers))
+        else:
+            cross_attn_layers = config.cross_attn_layer
+
+        self.layers = nn.ModuleList([PythiaSeekLayer(config, add_cross_attention=i in cross_attn_layers) for i in range(config.num_hidden_layers)])
+        self.final_layer_norm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+
+        self.gradient_checkpointing = False
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def get_input_embeddings(self):
+        return self.embed_in
+
+    def set_input_embeddings(self, value):
+        self.embed_in = value
+
+    def forward(
+        self,
+        input_ids: Optional[torch.LongTensor] = None,
+        encoder_hidden_states: Optional[torch.FloatTensor] = None,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        step: Optional[int] = None
+    ) -> Union[Tuple, BaseModelOutputWithPast]:
+        r"""
+        past_key_values (`tuple(tuple(torch.FloatTensor))` of length `config.n_layers` with each tuple having 4 tensors of shape `(batch_size, num_heads, sequence_length - 1, embed_size_per_head)`):
+            Contains precomputed key and value hidden states of the attention blocks. Can be used to speed up decoding.
+            If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
+            don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
+            `decoder_input_ids` of shape `(batch_size, sequence_length)`.
+        use_cache (`bool`, *optional*):
+            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
+            `past_key_values`).
+        """
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
+
+        if input_ids is not None and inputs_embeds is not None:
+            raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
+        elif input_ids is not None:
+            input_shape = input_ids.size()
+        elif inputs_embeds is not None:
+            input_shape = inputs_embeds.size()[:-1]
+        else:
+            raise ValueError("You have to specify either input_ids or inputs_embeds")
+
+        batch_size, seq_length = input_shape
+
+        if past_key_values is None:
+            past_length = 0
+            past_key_values = tuple([None] * self.config.num_hidden_layers)
+        else:
+            past_length = past_key_values[0][0].size(-2)
+
+        if position_ids is None:
+            device = input_ids.device if input_ids is not None else inputs_embeds.device
+            position_ids = torch.arange(past_length, seq_length + past_length, dtype=torch.long, device=device)
+            position_ids = position_ids.unsqueeze(0).view(-1, seq_length)
+        else:
+            position_ids = position_ids.view(-1, seq_length).long()
+
+        # Attention mask.
+        if attention_mask is not None:
+            assert batch_size > 0, "batch_size has to be defined and > 0"
+            attention_mask = attention_mask.view(batch_size, -1)
+            # We create a 3D attention mask from a 2D tensor mask.
+            # Sizes are [batch_size, 1, 1, to_seq_length]
+            # So we can broadcast to [batch_size, num_heads, from_seq_length, to_seq_length]
+            # this attention mask is more simple than the triangular masking of causal attention
+            # used in OpenAI GPT, we just need to prepare the broadcast dimension here.
+            attention_mask = attention_mask[:, None, None, :]
+
+            # Since attention_mask is 1.0 for positions we want to attend and 0.0 for
+            # masked positions, this operation will create a tensor which is 0.0 for
+            # positions we want to attend and the dtype's smallest value for masked positions.
+            # Since we are adding it to the raw scores before the softmax, this is
+            # effectively the same as removing these entirely.
+            attention_mask = attention_mask.to(dtype=self.dtype)  # fp16 compatibility
+            attention_mask = (1.0 - attention_mask) * torch.finfo(self.dtype).min
+
+        # Prepare head mask if needed
+        # 1.0 in head_mask indicate we keep the head
+        # attention_probs has shape bsz x n_heads x N x N
+        # input head_mask has shape [num_heads] or [num_hidden_layers x num_heads]
+        # and head_mask is converted to shape [num_hidden_layers x batch x num_heads x seq_length x seq_length]
+        head_mask = self.get_head_mask(head_mask, self.config.num_hidden_layers)
+
+        if inputs_embeds is None:
+            inputs_embeds = self.embed_in(input_ids)
+
+        hidden_states = inputs_embeds
+
+        if self.gradient_checkpointing and self.training:
+            if use_cache:
+                logger.warning(
+                    "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..."
+                )
+                use_cache = False
+
+        presents = () if use_cache else None
+        all_attentions = () if output_attentions else None
+        all_hidden_states = () if output_hidden_states else None
+        for i, (layer, layer_past) in enumerate(zip(self.layers, past_key_values)):
+            if output_hidden_states:
+                all_hidden_states = all_hidden_states + (hidden_states,)
+
+            if self.gradient_checkpointing and self.training:
+
+                def create_custom_forward(module):
+                    def custom_forward(*inputs):
+                        # None for layer_past
+                        return module(*inputs, use_cache, None, output_attentions, step)
+
+                    return custom_forward
+
+                outputs = torch.utils.checkpoint.checkpoint(
+                    create_custom_forward(layer),
+                    hidden_states,
+                    encoder_hidden_states,
+                    attention_mask,
+                    position_ids,
+                    head_mask[i],
+                )
+            else:
+                outputs = layer(
+                    hidden_states=hidden_states,
+                    encoder_hidden_states=encoder_hidden_states,
+                    attention_mask=attention_mask,
+                    position_ids=position_ids,
+                    head_mask=head_mask[i],
+                    layer_past=layer_past,
+                    use_cache=use_cache,
+                    output_attentions=output_attentions,
+                    step=step,
+                )
+            hidden_states = outputs[0]
+            if use_cache is True:
+                presents = presents + (outputs[1],)
+            if output_attentions:
+                all_attentions = all_attentions + (outputs[2 if use_cache else 1],)
+
+        hidden_states = self.final_layer_norm(hidden_states)
+        # Add last hidden state
+        if output_hidden_states:
+            all_hidden_states = all_hidden_states + (hidden_states,)
+
+        if not return_dict:
+            return tuple(v for v in [hidden_states, presents, all_hidden_states, all_attentions] if v is not None)
+
+        return BaseModelOutputWithPast(
+            last_hidden_state=hidden_states,
+            past_key_values=presents,
+            hidden_states=all_hidden_states,
+            attentions=all_attentions,
+        )
+
+
+class PythiaSeekForCausalLM(PythiaSeekPreTrainedModel):
+    _keys_to_ignore_on_load_missing = [r"position_ids", r"predictions.decoder.bias"]
+
+    def __init__(self, config):
+        super().__init__(config)
+
+        self.gpt_neox = PythiaSeekModel(config)
+        self.embed_out = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
+
+        self.hidden_size = config.hidden_size
+        self.encoder_dim = config.encoder_dim
+
+        if self.hidden_size != self.encoder_dim:
+            self.enc_dec_proj = nn.Sequential(nn.Linear(config.encoder_dim, config.hidden_size * 4),
+                                              nn.ReLU(),
+                                              nn.Linear(config.hidden_size * 4, config.hidden_size),
+                                              )
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def get_output_embeddings(self):
+        return self.embed_out
+
+    def set_output_embeddings(self, new_embeddings):
+        self.embed_out = new_embeddings
+
+    def forward(
+        self,
+        input_ids: torch.LongTensor,
+        encoder_hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        step: Optional[int] = None
+    ) -> Union[Tuple, CausalLMOutputWithPast]:
+        r"""
+        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
+            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
+            `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of shape
+            `(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`. The two additional tensors are
+            only required when the model is used as a decoder in a Sequence to Sequence model.
+
+            Contains pre-computed hidden-states (key and values in the self-attention blocks that can be used (see
+            `past_key_values` input) to speed up sequential decoding.
+
+            If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
+            don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
+            `decoder_input_ids` of shape `(batch_size, sequence_length)`.
+        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+            Labels for computing the left-to-right language modeling loss (next word prediction). Indices should be in
+            `[-100, 0, ..., config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are
+            ignored (masked), the loss is only computed for the tokens with labels n `[0, ..., config.vocab_size]`.
+        use_cache (`bool`, *optional*):
+            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
+            `past_key_values`).
+
+        Returns:
+
+        Example:
+
+        ```python
+        >>> from transformers import AutoTokenizer, PythiaSeekForCausalLM, PythiaSeekConfig
+        >>> import torch
+
+        >>> tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neox-20b")
+        >>> config = PythiaSeekConfig.from_pretrained("EleutherAI/gpt-neox-20b")
+        >>> config.is_decoder = True
+        >>> model = PythiaSeekForCausalLM.from_pretrained("EleutherAI/gpt-neox-20b", config=config)
+
+        >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
+        >>> outputs = model(**inputs)
+
+        >>> prediction_logits = outputs.logits
+        ```"""
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        if self.hidden_size != self.encoder_dim:
+            encoder_hidden_states = encoder_hidden_states.to(self.enc_dec_proj[0].weight.dtype)
+            encoder_hidden_states = self.enc_dec_proj(encoder_hidden_states)
+
+        outputs = self.gpt_neox(
+            input_ids,
+            encoder_hidden_states=encoder_hidden_states,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            past_key_values=past_key_values,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            step=step,
+        )
+
+        hidden_states = outputs[0]
+        lm_logits = self.embed_out(hidden_states)
+
+        lm_loss = None
+        if labels is not None:
+            # move labels to correct device to enable model parallelism
+            labels = labels.to(lm_logits.device)
+            # we are doing next-token prediction; shift prediction scores and input ids by one
+            shift_logits = lm_logits[:, :-1, :].contiguous()
+            labels = labels[:, 1:].contiguous()
+            loss_fct = CrossEntropyLoss()
+            lm_loss = loss_fct(shift_logits.view(-1, shift_logits.size(-1)), labels.view(-1))
+
+        if not return_dict:
+            output = (lm_logits,) + outputs[1:]
+            return ((lm_loss,) + output) if lm_loss is not None else output
+
+        return CausalLMOutputWithPast(
+            loss=lm_loss,
+            logits=lm_logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+        )
+
+    def prepare_inputs_for_generation(
+        self, input_ids, past_key_values=None, attention_mask=None, inputs_embeds=None, **kwargs
+    ):
+        input_shape = input_ids.shape
+
+        # cut decoder_input_ids if past is used
+        if past_key_values and past_key_values[0] is not None:
+            input_ids = input_ids[:, -1:]
+
+        position_ids = kwargs.get("position_ids", None)
+        if attention_mask is not None and position_ids is None:
+            # create position_ids on the fly for batch generation
+            position_ids = attention_mask.long().cumsum(-1) - 1
+            position_ids.masked_fill_(attention_mask == 0, 1)
+            if past_key_values:
+                position_ids = position_ids[:, -1].unsqueeze(-1)
+
+        # if model is used as a decoder in encoder-decoder model, the decoder attention mask is created on the fly
+        if attention_mask is None:
+            attention_mask = input_ids.new_ones(input_shape)
+
+        # if `inputs_embeds` are passed, we only want to use them in the 1st generation step
+        if inputs_embeds is not None and past_key_values is None:
+            model_inputs = {"inputs_embeds": inputs_embeds}
+        else:
+            model_inputs = {"input_ids": input_ids}
+
+        model_inputs.update(
+            {
+                "attention_mask": attention_mask,
+                "past_key_values": past_key_values,
+                "position_ids": position_ids,
+            }
+        )
+
+        return model_inputs
+
+    def _reorder_cache(self, past_key_values, beam_idx):
+        reordered_past = ()
+        for layer_past in past_key_values:
+            reordered_past += (
+                tuple(past_state.index_select(0, beam_idx) for past_state in layer_past[:2]) + layer_past[2:],
+            )
+        return reordered_past

gpt4all/models/pythiaseek/test_pythiaseek.py

@@ -0,0 +1,55 @@
+import torch
+from gpt4all.models import PythiaSeekConfig, PythiaSeekForCausalLM
+from transformers import AutoTokenizer, AutoModel
+
+# seed torch 
+
+torch.manual_seed(0)
+
+config = PythiaSeekConfig(encoder_dim=384, 
+                          hidden_size=1024, 
+                          intermediate_size=4096,
+                          num_hidden_layers=4,
+                          num_attention_heads=4)
+print("loaded config")
+
+print("loading model")
+model = PythiaSeekForCausalLM(config)
+print("loaded model")
+
+
+tokenizer =  AutoTokenizer.from_pretrained("EleutherAI/pythia-1b")
+tokenizer.pad_token = tokenizer.eos_token
+tokenizer.model_max_length = 1024
+
+encoder_tokenizer = AutoTokenizer.from_pretrained('sentence-transformers/all-MiniLM-L6-v2')
+encoder = AutoModel.from_pretrained('sentence-transformers/all-MiniLM-L6-v2')
+
+
+def mean_pooling(model_output, attention_mask):
+    token_embeddings = model_output[0] #First element of model_output contains all token embeddings
+    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
+    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
+
+text = "The quick brown fox jumps over the lazy dog."
+print("Encoded knn")
+tokenized = encoder_tokenizer(text, return_tensors="pt")
+
+# bs, seq_len, dim
+encodings = mean_pooling(encoder(**tokenized), tokenized["attention_mask"])
+
+# make 2 neighbors
+# (bs, knn, encoding_dim)
+encoder_hidden_states = torch.stack([encodings, encodings]).squeeze().unsqueeze(0)
+
+inputs = "What did the fox do?"
+
+print("Encoded inputs")
+tokenized_input = tokenizer([inputs], padding="max_length", truncation=True, return_tensors="pt")
+
+print("Running model")
+outputs = model(**tokenized_input, encoder_hidden_states=encoder_hidden_states)
+
+print(outputs)
+print(outputs[0].shape)
+

gpt4all/train/metrics.py

@@ -0,0 +1,50 @@
+from collections import Counter
+import string 
+import re
+
+
+# adapted from huggingface
+def normalize_answer(s):
+    """Lower text and remove punctuation, articles and extra whitespace."""
+
+    def remove_articles(text):
+        return re.sub(r"\b(a|an|the)\b", " ", text)
+
+    def white_space_fix(text):
+        return " ".join(text.split())
+
+    def remove_punc(text):
+        exclude = set(string.punctuation)
+        return "".join(ch for ch in text if ch not in exclude)
+
+    def lower(text):
+        return text.lower()
+
+    return white_space_fix(remove_articles(remove_punc(lower(s))))
+
+
+def f1_score(predictions, ground_truths):
+    total_f1 = []
+    for prediction, ground_truth in zip(predictions, ground_truths):
+        prediction_tokens = normalize_answer(prediction).split()
+        ground_truth_tokens = normalize_answer(ground_truth).split()
+        common = Counter(prediction_tokens) & Counter(ground_truth_tokens)
+        num_same = sum(common.values())
+        if num_same == 0:
+            total_f1.append(0)
+            continue
+
+        precision = 1.0 * num_same / len(prediction_tokens)
+        recall = 1.0 * num_same / len(ground_truth_tokens)
+        f1 = (2 * precision * recall) / (precision + recall)
+        total_f1.append(f1)
+
+    return total_f1
+
+
+def exact_match_score(predictions, ground_truths):
+    exact_scores = []
+    for prediction, ground_truth in zip(predictions, ground_truths):
+        exact_scores.append(normalize_answer(prediction) == normalize_answer(ground_truth))
+
+    return exact_scores

gpt4all/train/train.py

@@ -1,13 +1,13 @@
 import os
-from transformers import AutoModelForCausalLM, AutoTokenizer, get_scheduler, LlamaForCausalLM
+from transformers import AutoModelForCausalLM, AutoTokenizer, get_scheduler
 import torch
 from torch.optim import AdamW
 from argparse import ArgumentParser
-from read import read_config
+from gpt4all.utils.read import read_config
 from accelerate import Accelerator
 from accelerate.utils import DummyScheduler, DummyOptim, set_seed
 from peft import get_peft_model, LoraConfig, TaskType
-from data import load_data
+from gpt4all.data.instruction_tuning_dataloader import load_data
 from torchmetrics import MeanMetric
 from tqdm import tqdm
 import wandb
@@ -104,7 +104,7 @@ def train(accelerator, config):
         )
     else:
         scheduler = DummyScheduler(
-            optimizer, total_num_steps=config["warmup_steps"], warmup_num_steps=config["warmup_steps"]
+            optimizer, total_num_steps=total_num_steps, warmup_num_steps=config["warmup_steps"]
         )
 
     model, optimizer, train_dataloader, val_dataloader, scheduler = accelerator.prepare(

gpt4all/train/train_mem_retrieval.py

@@ -0,0 +1,328 @@
+import os
+import torch.nn.functional as F
+from transformers import AutoTokenizer, get_scheduler, AutoConfig
+import torch
+from torch.optim import AdamW
+from argparse import ArgumentParser
+from gpt4all.utils.read import read_config
+from accelerate import Accelerator
+from accelerate.utils import DummyScheduler, DummyOptim, set_seed
+from gpt4all.data.retrieval_dataloader import load_memory_augmented_data
+from torchmetrics import MeanMetric
+from tqdm import tqdm
+from gpt4all.models import LetheForCausalLM
+from gpt4all.models.lethe.modeling_lethe import MemoryIndex
+import wandb
+import pyarrow as pa
+from pyarrow import feather
+
+torch.backends.cuda.matmul.allow_tf32 = True
+
+def format_metrics(metrics, split, prefix=""):
+    log = f"[{split}]" + prefix
+    log += " ".join([f"{key}: {value:.4f}" for key, value in metrics.items()])
+
+    return log
+
+
+def calculate_per_example_loss(logits, labels):
+    lm_logits = logits[:, :-1, :].contiguous()
+    lm_labels = labels[:, 1:].contiguous()
+    loss = F.cross_entropy(lm_logits.view(-1, lm_logits.size(-1)), lm_labels.view(-1), reduction="none")
+    loss = loss.reshape(labels.shape[0], -1).mean(dim=-1)
+
+    # return tensor of shape (B,) where B is the batch size
+    return loss.cpu().tolist()
+
+
+def evaluate(model, index, config, val_dataloader, main_process=False):
+    model.eval()
+    val_loss = MeanMetric(nan_strategy="error").to(model.device)
+
+    ids = []
+    losses = []
+    with torch.no_grad():
+        for batch in tqdm(val_dataloader, disable=not main_process):
+            batch["id"] = batch["id"].detach().cpu()
+            memories = batch["retrieved_context"]
+
+            memories = memories[:, :config["num_neighbors_to_store"], :]
+            memories = memories.reshape(-1, memories.shape[-1])
+
+            for chunk_start in range(0, memories.shape[0], config["mem_chunk_size"]):
+                chunk_end = min(memories.shape[0], chunk_start + config["mem_chunk_size"])
+                mem_chunk = memories[chunk_start:chunk_end]
+                model(input_ids=mem_chunk)
+
+            qa_inputs = batch["input_ids"]
+            qa_labels = batch["labels"]
+            outputs = model(input_ids=qa_inputs, 
+                            labels=qa_labels, 
+            )
+
+            del memories
+            torch.cuda.empty_cache()
+            loss_values = accelerator.gather_for_metrics({"loss": outputs["loss"].detach()})
+
+            index.reset()
+            val_loss.update(loss_values["loss"])
+
+            per_example_loss = calculate_per_example_loss(outputs["logits"], qa_labels)
+
+            losses.extend(per_example_loss)
+            ids.extend(batch["id"].tolist())
+
+    ids = pa.array(ids)
+    losses = pa.array(losses)
+    schema = pa.schema([("loss", pa.float64()), ("id", pa.int32())])
+    table = pa.Table.from_arrays([losses, ids], schema=schema)
+    return val_loss, table
+
+
+def train(accelerator, config):
+    set_seed(config['seed'])
+
+    accelerator.print(config)
+    accelerator.print(f"Using {accelerator.num_processes} GPUs")
+
+    tokenizer = AutoTokenizer.from_pretrained(config['tokenizer_name'], model_max_length=config['max_length'])
+    # if no pad token, set it to eos
+    if tokenizer.pad_token is None:
+        tokenizer.pad_token = tokenizer.eos_token
+
+        
+    with accelerator.main_process_first():
+        train_dataloader, val_dataloader = load_memory_augmented_data(config, tokenizer) 
+
+
+    if accelerator.state.deepspeed_plugin is not None:
+        gradient_accumulation_steps = accelerator.state.deepspeed_plugin.deepspeed_config[
+            "gradient_accumulation_steps"
+        ]
+
+    accelerator.print(f"Len of train_dataloader: {len(train_dataloader)}")
+    total_num_steps = (len(train_dataloader) / gradient_accumulation_steps) * config["num_epochs"]
+    # instead of decaying to zero, decay to ratio of min_lr / lr
+    accelerator.print(f"Total training steps: {total_num_steps}")
+
+    checkpoint = config["gradient_checkpointing"]
+    
+    model_config = AutoConfig.from_pretrained(config["model_name"])
+
+    head_size = model_config.hidden_size // model_config.num_attention_heads
+    index = MemoryIndex(head_size,
+                        config["num_memories_per_index"],
+                        model_config.num_attention_heads
+    )
+    model = LetheForCausalLM.from_pretrained(config["model_name"], 
+                                             revision=config['version'] if 'version' in config else None,
+                                             use_cache=False if checkpoint else True,
+                                             memory_attn_layer=config["memory_attn_layer"],
+                                             num_neighbors_to_retrieve=config["num_neighbors_to_retrieve"],
+                                             index=index,
+                                             tracker=accelerator.get_tracker("wandb"),
+                                            )
+
+
+    accelerator.print(f"Training a {model.num_parameters():,} parameter model")
+    if checkpoint:
+        model.gradient_checkpointing_enable()
+
+    optimizer_cls = (
+        AdamW
+        if accelerator.state.deepspeed_plugin is None
+        or "optimizer" not in accelerator.state.deepspeed_plugin.deepspeed_config
+        else DummyOptim
+    )
+
+    # karpathy doesn't decay embeddding, maybe we should exclude
+    # https://github.com/karpathy/minGPT/commit/bbbdac74fa9b2e55574d70056163ffbae42310c1#diff-2075fa9c224b395be5bda85544dd36572b59c76c54562819eadadbf268602834R157s
+    optimizer = optimizer_cls(model.parameters(), lr=config["lr"], weight_decay=config["weight_decay"])
+
+
+    # Creates Dummy Scheduler if `scheduler` was spcified in the config file else creates `args.lr_scheduler_type` Scheduler
+    if config["scheduler"] or "scheduler" in accelerator.state.deepspeed_plugin.deepspeed_config:
+        if (
+            accelerator.state.deepspeed_plugin is None
+            or "scheduler" not in accelerator.state.deepspeed_plugin.deepspeed_config
+        ):
+            scheduler = get_scheduler(
+                name="cosine",
+                optimizer=optimizer,
+                num_warmup_steps=config["warmup_steps"] * accelerator.num_processes,
+                num_training_steps=total_num_steps,
+            )
+        else:
+            scheduler = DummyScheduler(
+                optimizer, total_num_steps=total_num_steps, warmup_num_steps=config["warmup_steps"]
+            )
+        model, optimizer, train_dataloader, val_dataloader, scheduler = accelerator.prepare(
+                model, optimizer, train_dataloader, val_dataloader, scheduler
+        )
+        use_scheduler = True
+    else:
+        model, optimizer, train_dataloader, val_dataloader = accelerator.prepare(
+                model, optimizer, train_dataloader, val_dataloader
+        )
+        use_scheduler = False
+
+    #  setup for saving training states in case preemption
+    if use_scheduler:
+        accelerator.register_for_checkpointing(scheduler)
+
+    if config["checkpoint"]:
+        accelerator.load_state(config["checkpoint"])
+        accelerator.print(f"Resumed from checkpoint: {config['checkpoint']}")
+        path = os.path.basename(config["train_args"]["resume_from_checkpoint"])
+        training_difference = os.path.splitext(path)[0]
+        resume_step = int(training_difference.replace("step_", ""))
+        accelerator.skip_first_batches(train_dataloader, resume_step)
+        accelerator.print(f"Resuming from step {resume_step}")
+
+    # log gradients
+    if accelerator.is_main_process and config["wandb"]:
+        wandb.watch(model, log_freq=config["log_grads_every"], log="all")
+
+    main_process = accelerator.is_main_process
+
+    for epoch in range(config["num_epochs"]):
+        train_loss = MeanMetric(nan_strategy="error").to(model.device)
+        for step, batch in enumerate(tqdm(train_dataloader, disable=not main_process)):
+            curr_step = step + epoch * len(train_dataloader)
+            memories = batch["retrieved_context"]
+            memories = memories[:, :config["num_neighbors_to_store"], :]
+            memories = memories.reshape(-1, memories.shape[-1])
+
+            # need to set to eval so we don't do mem attn as it's slow
+            model.eval()
+            with torch.no_grad():
+                for chunk_start in range(0, memories.shape[0], config["mem_chunk_size"]):
+                    chunk_end = min(memories.shape[0], chunk_start + config["mem_chunk_size"])
+                    mem_chunk = memories[chunk_start:chunk_end]
+                    model(input_ids=mem_chunk)
+
+            del memories
+            torch.cuda.empty_cache()
+
+            model.train()
+            qa_inputs = batch["input_ids"]
+            qa_labels = batch["labels"]
+            outputs = model(input_ids=qa_inputs, 
+                            labels=qa_labels, 
+                            log_attn_scores=True,
+                            step=curr_step,
+                            save_kv=False,
+            )
+            loss = outputs.loss
+            if config["wandb"]:
+                accelerator.log({"loss": loss}, step=curr_step)
+
+
+            # gather loss before backprop in case of gradient accumulation
+            loss_values = accelerator.gather_for_metrics({"loss": loss.detach().float()})
+            train_loss.update(loss_values["loss"])
+
+            loss = loss / gradient_accumulation_steps
+            accelerator.backward(loss)
+            # !! don't reset index until after backwards pass
+            index.reset()
+            # get gradient norm of all params
+
+            # log LR in case something weird happens 
+            if config["wandb"]:
+                if step > 0 and step % (config["log_lr_every"] ) == 0:
+                    lr = optimizer.param_groups[0]["lr"]
+                    accelerator.log({"lr": lr}, step=curr_step)
+
+            if (step + 1) % gradient_accumulation_steps == 0 or step == len(train_dataloader) - 1:
+                optimizer.step()
+                if use_scheduler:
+                    scheduler.step()
+                optimizer.zero_grad()
+
+            if step > 0 and config["save_every"] > 0 and step % config["save_every"] == 0:
+                # accelerator.save_state(f"{config['output_dir']}/step_{curr_step}")
+                unwrapped_model = accelerator.unwrap_model(model)
+                unwrapped_model.save_pretrained(
+                        f"{config['output_dir']}/step_{step}",
+                        is_main_process=accelerator.is_main_process,
+                        save_function=accelerator.save,
+                        state_dict=accelerator.get_state_dict(model),
+                )
+
+            if step > 0 and (step % config["eval_every"] == 0 or step == len(train_dataloader) - 1):
+                val_loss, loss_table = evaluate(model, index, config, val_dataloader, main_process=main_process)
+
+                local_rank = accelerator.process_index
+                feather.write_feather(loss_table, f"{config['output_dir']}/val_losses_step_{curr_step}_rank_{local_rank}.feather")
+
+                log_train = {
+                        "train_loss": train_loss.compute()
+                    }
+                log_val = {
+                    "val_loss": val_loss.compute(),
+                }
+
+                if config["wandb"]:
+                    curr_step = step + epoch * len(train_dataloader)
+                    accelerator.log({**log_train, **log_val}, step=curr_step)
+
+                accelerator.print(f"Current LR: {optimizer.param_groups[0]['lr']}")
+                accelerator.print(format_metrics(log_train, "train", f" step {step} "))
+                accelerator.print(format_metrics(log_val, "val", f" step {step} "))
+
+                train_loss.reset()
+
+        accelerator.print(f"Epoch {epoch} finished")
+        accelerator.wait_for_everyone()
+        unwrapped_model = accelerator.unwrap_model(model)
+        if config["push_to_hub"]:
+            accelerator.print(f"Pushing to HF hub")
+            try:
+                if accelerator.is_main_process:
+                    unwrapped_model.push_to_hub(config["save_name"] + f"-epoch_{epoch}", private=True)
+
+            except Exception as e:
+                accelerator.print(e)
+                accelerator.print(f"Failed to push to hub")
+
+        unwrapped_model.save_pretrained(
+            f"{config['output_dir']}/epoch_{epoch}",
+            is_main_process=accelerator.is_main_process,
+            save_function=accelerator.save,
+            state_dict=accelerator.get_state_dict(model),
+        )
+            
+    accelerator.wait_for_everyone()
+    unwrapped_model = accelerator.unwrap_model(model)
+    unwrapped_model.save_pretrained(
+        f"{config['output_dir']}/final",
+        is_main_process=accelerator.is_main_process,
+        save_function=accelerator.save,
+        state_dict=accelerator.get_state_dict(model),
+    )
+
+    accelerator.end_training()
+
+    
+
+if __name__ == "__main__":
+    # parse arguments by reading in a config
+    parser = ArgumentParser()
+    parser.add_argument("--config", type=str, default="config.yaml")
+
+    args = parser.parse_args()
+
+    config = read_config(args.config)
+
+    if config["wandb"]:
+        accelerator = Accelerator(log_with="wandb")
+        accelerator.init_trackers(
+            project_name=config["wandb_project_name"],
+            config=config,
+            init_kwargs={"wandb": {"entity": config["wandb_entity"]}},
+        )
+    else:
+        accelerator = Accelerator()
+
+    train(accelerator, config=config)

gpt4all/train/train_retrieval.py

@@ -0,0 +1,307 @@
+import os
+from transformers import AutoModelForCausalLM, AutoTokenizer, get_scheduler
+import torch
+from torch.optim import AdamW
+from argparse import ArgumentParser
+from gpt4all.utils.read import read_config
+from accelerate import Accelerator
+from accelerate.utils import DummyScheduler, DummyOptim, set_seed
+from peft import get_peft_model, LoraConfig, TaskType
+from gpt4all.data.retrieval_dataloader import load_retrieval_augmented_data
+from torchmetrics import MeanMetric
+from tqdm import tqdm
+from gpt4all.models import GPTJRForCausalLM, PythiaSeekForCausalLM
+import wandb
+
+torch.backends.cuda.matmul.allow_tf32 = True
+
+def format_metrics(metrics, split, prefix=""):
+    log = f"[{split}]" + prefix
+    log += " ".join([f"{key}: {value:.4f}" for key, value in metrics.items()])
+
+    return log
+
+
+def evaluate(model, val_dataloader, step, main_process=False):
+    model.eval()
+    val_loss = MeanMetric(nan_strategy="error").to(model.device)
+
+    with torch.no_grad():
+        for batch in tqdm(val_dataloader, disable=not main_process):
+            outputs = model(input_ids=batch["input_ids"], 
+                            labels=batch["labels"], 
+                            encoder_hidden_states=batch["encoder_hidden_states"],
+                            step=step)
+            loss_values = accelerator.gather_for_metrics({"loss": outputs["loss"].detach()})
+
+
+            val_loss.update(loss_values["loss"])
+
+    return val_loss
+
+
+def train(accelerator, config):
+    set_seed(config['seed'])
+
+    accelerator.print(config)
+    accelerator.print(f"Using {accelerator.num_processes} GPUs")
+
+    tokenizer = AutoTokenizer.from_pretrained(config['tokenizer_name'], model_max_length=config['max_length'])
+    # if no pad token, set it to eos
+    if tokenizer.pad_token is None:
+        tokenizer.pad_token = tokenizer.eos_token
+
+        
+    with accelerator.main_process_first():
+        train_dataloader, val_dataloader = load_retrieval_augmented_data(config, tokenizer) 
+
+    if accelerator.state.deepspeed_plugin is not None:
+        gradient_accumulation_steps = accelerator.state.deepspeed_plugin.deepspeed_config[
+            "gradient_accumulation_steps"
+        ]
+
+    accelerator.print(f"Len of train_dataloader: {len(train_dataloader)}")
+    total_num_steps = (len(train_dataloader) / gradient_accumulation_steps) * config["num_epochs"]
+    # instead of decaying to zero, decay to ratio of min_lr / lr
+    accelerator.print(f"Total training steps: {total_num_steps}")
+
+    checkpoint = config["gradient_checkpointing"]
+    #ensures back compat with non retrieval models
+    if 'encoder_dim' in config:
+        if "gptj" in config["model_name"]:
+            model = GPTJRForCausalLM.from_pretrained(config["model_name"], 
+                                                    revision=config['version'] if 'version' in config else None,
+                                                    use_cache=False if checkpoint else True,
+                                                    encoder_dim=config["encoder_dim"],
+                                                    # hardcoded!! TODO: change to config
+                                                    total_alpha_steps=10000,
+                                                    ) 
+        elif "pythia" in config["model_name"]:
+            cross_attn_layer = config["cross_attn_layer"]
+            learnable_alpha = config["learnable_alpha"]
+            model, info = PythiaSeekForCausalLM.from_pretrained(config["model_name"], 
+                                                    revision=config['version'] if 'version' in config else None,
+                                                    use_cache=False if checkpoint else True,
+                                                    encoder_dim=config["encoder_dim"],
+                                                    # hardcoded!! TODO: change to config
+                                                    total_alpha_steps=2500,
+                                                    # mem transformer uses 9/12, pythia1-b is 16 layers
+                                                    cross_attn_layer=cross_attn_layer,
+                                                    learnable_alpha=learnable_alpha,
+                                                    output_loading_info=True
+                                                    )
+            # freeze all pretrained layers
+            if config["freeze_pretrained"]:
+                for name, param in model.named_parameters():
+                    if name not in info["missing_keys"]:
+                        param.requires_grad = False
+
+    else:
+        model = AutoModelForCausalLM.from_pretrained(config["model_name"], 
+                                                     use_cache=False if checkpoint else True,
+                                                     trust_remote_code=True) 
+
+    accelerator.print(f"Training a {model.num_parameters():,} parameter model")
+    if checkpoint:
+        model.gradient_checkpointing_enable()
+
+    if config["lora"]:
+        peft_config = LoraConfig(
+            # should R be configurable?
+            task_type=TaskType.CAUSAL_LM, inference_mode=False, r=8, lora_alpha=32, lora_dropout=0.1
+        )
+        model = get_peft_model(model, peft_config)
+        model.print_trainable_parameters()
+
+    optimizer_cls = (
+        AdamW
+        if accelerator.state.deepspeed_plugin is None
+        or "optimizer" not in accelerator.state.deepspeed_plugin.deepspeed_config
+        else DummyOptim
+    )
+
+    # karpathy doesn't decay embeddding, maybe we should exclude
+    # https://github.com/karpathy/minGPT/commit/bbbdac74fa9b2e55574d70056163ffbae42310c1#diff-2075fa9c224b395be5bda85544dd36572b59c76c54562819eadadbf268602834R157s
+    optimizer = optimizer_cls(model.parameters(), lr=config["lr"], weight_decay=config["weight_decay"])
+
+
+    # Creates Dummy Scheduler if `scheduler` was spcified in the config file else creates `args.lr_scheduler_type` Scheduler
+    if config["scheduler"] or "scheduler" in accelerator.state.deepspeed_plugin.deepspeed_config:
+        if (
+            accelerator.state.deepspeed_plugin is None
+            or "scheduler" not in accelerator.state.deepspeed_plugin.deepspeed_config
+        ):
+            scheduler = get_scheduler(
+                name="cosine",
+                optimizer=optimizer,
+                num_warmup_steps=config["warmup_steps"] * accelerator.num_processes,
+                num_training_steps=total_num_steps,
+            )
+        else:
+            scheduler = DummyScheduler(
+                optimizer, total_num_steps=total_num_steps, warmup_num_steps=config["warmup_steps"]
+            )
+        model, optimizer, train_dataloader, val_dataloader, scheduler = accelerator.prepare(
+                model, optimizer, train_dataloader, val_dataloader, scheduler
+        )
+        scheduler = True
+    else:
+        model, optimizer, train_dataloader, val_dataloader = accelerator.prepare(
+                model, optimizer, train_dataloader, val_dataloader
+        )
+        scheduler = False
+
+
+    #  setup for saving training states in case preemption
+    if scheduler:
+        accelerator.register_for_checkpointing(scheduler)
+
+    if config["checkpoint"]:
+        accelerator.load_state(config["checkpoint"])
+        accelerator.print(f"Resumed from checkpoint: {config['checkpoint']}")
+        path = os.path.basename(config["train_args"]["resume_from_checkpoint"])
+        training_difference = os.path.splitext(path)[0]
+        resume_step = int(training_difference.replace("step_", ""))
+        accelerator.skip_first_batches(train_dataloader, resume_step)
+        accelerator.print(f"Resuming from step {resume_step}")
+
+
+    # log gradients
+    if accelerator.is_main_process and config["wandb"]:
+        wandb.watch(model, log_freq=config["log_grads_every"], log="all")
+
+    main_process = accelerator.is_main_process
+
+    if learnable_alpha is False and isinstance(cross_attn_layer, int):
+        accelerator.log({"alpha": model.gpt_neox.layers[cross_attn_layer]._update_alpha(0)}, step=0)
+
+    for epoch in range(config["num_epochs"]):
+        train_loss = MeanMetric(nan_strategy="error").to(model.device)
+        for step, batch in enumerate(tqdm(train_dataloader, disable=not main_process)):
+            model.train()
+            curr_step = step + len(train_dataloader) * epoch
+            outputs = model(input_ids=batch["input_ids"], 
+                            labels=batch["labels"], 
+                            encoder_hidden_states=batch["encoder_hidden_states"],
+                            step=curr_step)
+            loss = outputs.loss
+
+            if config["debug"]:
+                logits = outputs.logits
+                pred_tokens = torch.argmax(logits, dim=-1)
+                labels = batch["labels"].clone()
+
+                for row in range(labels.shape[0]):
+                    curr_label = labels[row]
+                    mask = curr_label != -100
+                    decoded_true = tokenizer.decode(curr_label[mask], skip_special_tokens=True)
+                    decoded_pred = tokenizer.decode(pred_tokens[row][mask], skip_special_tokens=True)
+
+                    accelerator.print(f"Predicted tokens: {decoded_pred}")
+                    accelerator.print(f"True tokens: {decoded_true}")
+
+
+            # gather loss before backprop in case of gradient accumulation
+            loss_values = accelerator.gather_for_metrics({"loss": loss.detach().float()})
+            train_loss.update(loss_values["loss"])
+
+            loss = loss / gradient_accumulation_steps
+            accelerator.backward(loss)
+            # get gradient norm of all params
+
+            # log LR in case something weird happens 
+            if config["wandb"]:
+                if step > 0 and step % (config["log_lr_every"] ) == 0:
+                    lr = optimizer.param_groups[0]["lr"]
+                    accelerator.log({"lr": lr}, step=curr_step)
+
+                    if learnable_alpha is False and isinstance(cross_attn_layer, int):
+                        curr_alpha = model.gpt_neox.layers[cross_attn_layer]._update_alpha(curr_step)
+                        accelerator.log({"alpha": curr_alpha}, step=curr_step)
+                    elif isinstance(cross_attn_layer, int):
+                        curr_alpha = model.gpt_neox.layers[cross_attn_layer].alpha.item()
+                        accelerator.log({"alpha": curr_alpha}, step=curr_step)
+
+            if (step + 1) % gradient_accumulation_steps == 0 or step == len(train_dataloader) - 1:
+                optimizer.step()
+                if scheduler:
+                    scheduler.step()
+                optimizer.zero_grad()
+
+
+            if step > 0 and config["save_every"] > 0 and step % config["save_every"] == 0:
+                accelerator.save_state(f"{config['output_dir']}/step_{curr_step}")
+
+            if step > 0 and (step % config["eval_every"] == 0 or step == len(train_dataloader) - 1):
+                val_loss = evaluate(model, val_dataloader, step=curr_step, main_process=main_process)
+
+                log_train = {
+                        "train_loss": train_loss.compute()
+                    }
+                log_val = {
+                    "val_loss": val_loss.compute(),
+                }
+
+                if config["wandb"]:
+                    curr_step = step + epoch * len(train_dataloader)
+                    accelerator.log({**log_train, **log_val}, step=curr_step)
+
+                accelerator.print(f"Current LR: {optimizer.param_groups[0]['lr']}")
+                accelerator.print(format_metrics(log_train, "train", f" step {step} "))
+                accelerator.print(format_metrics(log_val, "val", f" step {step} "))
+
+                train_loss.reset()
+
+        accelerator.print(f"Epoch {epoch} finished")
+        accelerator.wait_for_everyone()
+        unwrapped_model = accelerator.unwrap_model(model)
+        if config["push_to_hub"]:
+            accelerator.print(f"Pushing to HF hub")
+            try:
+                if accelerator.is_main_process:
+                    unwrapped_model.push_to_hub(config["save_name"] + f"-epoch_{epoch}", private=True)
+
+            except Exception as e:
+                accelerator.print(e)
+                accelerator.print(f"Failed to push to hub")
+
+        unwrapped_model.save_pretrained(
+            f"{config['output_dir']}/epoch_{epoch}",
+            is_main_process=accelerator.is_main_process,
+            save_function=accelerator.save,
+            state_dict=accelerator.get_state_dict(model),
+        )
+            
+    accelerator.wait_for_everyone()
+    unwrapped_model = accelerator.unwrap_model(model)
+    unwrapped_model.save_pretrained(
+        f"{config['output_dir']}/final",
+        is_main_process=accelerator.is_main_process,
+        save_function=accelerator.save,
+        state_dict=accelerator.get_state_dict(model),
+    )
+
+    accelerator.end_training()
+
+    
+
+if __name__ == "__main__":
+    # parse arguments by reading in a config
+    parser = ArgumentParser()
+    parser.add_argument("--config", type=str, default="config.yaml")
+
+    args = parser.parse_args()
+
+    config = read_config(args.config)
+
+    if config["wandb"]:
+        accelerator = Accelerator(log_with="wandb")
+        accelerator.init_trackers(
+            project_name=config["wandb_project_name"],
+            config=config,
+            init_kwargs={"wandb": {"entity": config["wandb_entity"]}},
+        )
+    else:
+        accelerator = Accelerator()
+
+    train(accelerator, config=config)

gpt4all/utils/distributed_utils.py

@@ -0,0 +1,27 @@
+import torch.distributed as dist
+from contextlib import contextmanager
+
+
+def rank0_print(msg):
+    if dist.is_initialized():
+        if dist.get_rank() == 0:
+            print(msg)
+    else:
+        print(msg)
+
+
+@contextmanager
+def main_process_first(is_main):
+    yield from _goes_first(is_main)
+    
+
+        
+def _goes_first(is_main):
+    if not is_main:
+        dist.barrier()
+
+    yield
+
+    if is_main:
+        dist.barrier()
+    

requirements.txt

@@ -12,4 +12,7 @@ sentencepiece
 jsonlines
 nomic
 scikit-learn
-matplotlib
+matplotlib
+apache_beam 
+mwparserfromhell
+hnswlib

setup.py

@@ -0,0 +1,34 @@
+from setuptools import setup, find_packages
+
+with open('README.md', 'r', encoding='utf-8') as f:
+    long_description = f.read()
+
+with open('requirements.txt', 'r', encoding='utf-8') as f:
+    requirements = [line.strip() for line in f if line.strip()]
+
+setup(
+    name='gpt4all',
+    version='0.0.1',
+    author='nomic-ai',
+    author_email='zach@nomic-ai',
+    description='an ecosystem of open-source chatbots trained on a massive collections of clean assistant data including code, stories and dialogue',
+    long_description=long_description,
+    long_description_content_type='text/markdown',
+    url='https://github.com/nomic-ai/gpt4all',
+    packages=find_packages(),
+    install_requires=requirements,
+    classifiers=[
+        'Development Status :: 3 - Alpha',
+        'License :: OSI Approved :: MIT License',
+        'Programming Language :: Python :: 3',
+        'Programming Language :: Python :: 3.6',
+        'Programming Language :: Python :: 3.7',
+        'Programming Language :: Python :: 3.8',
+        'Programming Language :: Python :: 3.9',
+        'Topic :: Text Processing :: Linguistic',
+        'Topic :: Scientific/Engineering :: Artificial Intelligence',
+        'Intended Audience :: Science/Research',
+        'Operating System :: OS Independent',
+    ],
+    python_requires='>=3.6',
+)