dollyrm14
1df0fceb4d
Merge 6fd3a23daf into b7ee43788d
2025-06-10 16:40:40 -05:00
google-labs-jules[bot]
6fd3a23daf
Refactor and enhance LLM prompt styles
...
This commit introduces several improvements to the prompt formatting logic in `private_gpt/components/llm/prompt_helper.py`:
1. **Llama3PromptStyle**:
* Implemented tool handling capabilities, allowing for the formatting of tool call and tool result messages within the Llama 3 prompt structure.
* Ensured correct usage of BOS, EOT, and other Llama 3 specific tokens.
2. **MistralPromptStyle**:
* Refactored the `_messages_to_prompt` method for more robust handling of various conversational scenarios, including consecutive user messages and initial assistant messages.
* Ensured correct application of `<s>`, `</s>`, and `[INST]` tags.
3. **ChatMLPromptStyle**:
* Corrected the logic for handling system messages to prevent duplication and ensure accurate ChatML formatting (`<|im_start|>role\ncontent<|im_end|>`).
4. **TagPromptStyle**:
* Addressed a FIXME comment by incorporating `<s>` (BOS) and `</s>` (EOS) tokens, making it more suitable for Llama-based models like Vigogne.
* Fixed a minor bug related to enum string conversion.
5. **Unit Tests**:
* Added a new test suite in `tests/components/llm/test_prompt_helper.py`.
* These tests provide comprehensive coverage for all modified prompt styles, verifying correct prompt generation for various inputs, edge cases, and special token placements.
These changes improve the correctness, robustness, and feature set of the supported prompt styles, leading to better compatibility and interaction with the respective language models.
2025-06-10 21:05:34 +00:00
Iván Martínez
b7ee43788d
Update README.md
2024-11-13 20:29:56 +01:00
meng-hui
940bdd49af
fix: 503 when private gpt gets ollama service ( #2104 )
...
When running private gpt with external ollama API, ollama service
returns 503 on startup because ollama service (traefik) might not be
ready.
- Add healthcheck to ollama service to test for connection to external
ollama
- private-gpt-ollama service depends on ollama being service_healthy
Co-authored-by: Koh Meng Hui <kohmh@duck.com>
2024-10-17 12:44:28 +02:00
Javier Martinez
5851b02378
feat: update llama-index + dependencies ( #2092 )
...
* chore: update libraries
* fix: mypy
* chore: more updates
* fix: mypy/black
* chore: fix docker warnings
* fix: mypy
* fix: black
2024-09-26 16:29:52 +02:00
Dmitri Qiu
5fbb402477
fix: Sanitize null bytes before ingestion ( #2090 )
...
* Sanitize null bytes before ingestion
* Added comments
2024-09-25 12:00:03 +02:00
J
fa3c30661d
fix: Add default mode option to settings ( #2078 )
...
* Add default mode option to settings
* Revise default_mode to Literal (enum) and add to settings.yaml
* Revise to pass make check/test
* Default mode: RAG
---------
Co-authored-by: Jason <jason@sowinsight.solutions>
2024-09-24 08:33:02 +02:00
Liam Dowd
f9182b3a86
feat: Adding MistralAI mode ( #2065 )
...
* Adding MistralAI mode
* Update embedding_component.py
* Update ui.py
* Update settings.py
* Update embedding_component.py
* Update settings.py
* Update settings.py
* Update settings-mistral.yaml
* Update llm_component.py
* Update settings-mistral.yaml
* Update settings.py
* Update settings.py
* Update ui.py
* Update embedding_component.py
* Delete settings-mistral.yaml
---------
Co-authored-by: SkiingIsFun123 <101684827+SkiingIsFun123@users.noreply.github.com>
Co-authored-by: Javier Martinez <javiermartinezalvarez98@gmail.com>
2024-09-24 08:31:30 +02:00
Javier Martinez
8c12c6830b
fix: docker permissions ( #2059 )
...
* fix: missing depends_on
* chore: update copy permissions
* chore: update entrypoint
* Revert "chore: update entrypoint"
This reverts commit f73a36af2ffabc3f66bb
2024-09-24 08:30:58 +02:00
Javier Martinez
77461b96cf
feat: add retry connection to ollama ( #2084 )
...
* feat: add retry connection to ollama
When Ollama is running in the docker-compose, traefik is not ready sometimes to route the request, and it fails
* fix: mypy
2024-09-16 16:43:05 +02:00
Trivikram Kamat
42628596b2
ci: bump actions/checkout to v4 ( #2077 )
2024-09-09 08:53:13 +02:00
Artur Martins
7603b3627d
fix: Rectify ffmpy poetry config; update version from 0.3.2 to 0.4.0 ( #2062 )
...
* Fix: Rectify ffmpy 0.3.2 poetry config
* keep optional set to false for ffmpy
* Updating ffmpy to version 0.4.0
* Remove comment about a fix
2024-08-21 10:39:58 +02:00
Javier Martinez
89477ea9d3
fix: naming image and ollama-cpu ( #2056 )
2024-08-12 08:23:16 +02:00
github-actions[bot]
22904ca8ad
chore(main): release 0.6.2 ( #2049 )
...
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
2024-08-08 18:16:41 +02:00
Javier Martinez
7fefe408b4
fix: auto-update version ( #2052 )
2024-08-08 16:50:42 +02:00
Javier Martinez
b1acf9dc2c
fix: publish image name ( #2043 )
2024-08-07 17:39:32 +02:00
Javier Martinez
4ca6d0cb55
fix: add numpy issue to troubleshooting ( #2048 )
...
* docs: add numpy issue to troubleshooting
* fix: troubleshooting link
...
2024-08-07 12:16:03 +02:00
Javier Martinez
b16abbefe4
fix: update matplotlib to 3.9.1-post1 to fix win install
...
* chore: block matplotlib to fix installation in window machines
* chore: remove workaround, just update poetry.lock
* fix: update matplotlib to last version
2024-08-07 11:26:42 +02:00
github-actions[bot]
ca2b8da69c
chore(main): release 0.6.1 ( #2041 )
...
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
2024-08-05 17:17:34 +02:00
Javier Martinez
f09f6dd255
fix: add built image from DockerHub ( #2042 )
...
* chore: update docker-compose with profiles
* docs: add quick start doc
* chore: generate docker release when new version is released
* chore: add dockerhub image in docker-compose
* docs: update quickstart with local/remote images
* chore: update docker tag
* chore: refactor dockerfile names
* chore: update docker-compose names
* docs: update llamacpp naming
* fix: naming
* docs: fix llamacpp command
2024-08-05 17:15:38 +02:00
Liam Dowd
1c665f7900
fix: Adding azopenai to model list ( #2035 )
...
Fixing the error I encountered while using the azopenai mode
2024-08-05 16:30:10 +02:00
Javier Martinez
1d4c14d7a3
fix(deploy): generate docker release when new version is released ( #2038 )
2024-08-05 16:28:19 +02:00
Javier Martinez
dae0727a1b
fix(deploy): improve Docker-Compose and quickstart on Docker ( #2037 )
...
* chore: update docker-compose with profiles
* docs: add quick start doc
2024-08-05 16:28:19 +02:00
github-actions[bot]
6674b46fea
chore(main): release 0.6.0 ( #1834 )
...
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
2024-08-02 11:28:22 +02:00
Javier Martinez
e44a7f5773
chore: bump version ( #2033 )
2024-08-02 11:26:03 +02:00
Javier Martinez
cf61bf780f
feat(llm): add progress bar when ollama is pulling models ( #2031 )
...
* fix: add ollama progress bar when pulling models
* feat: add ollama queue
* fix: mypy
2024-08-01 19:14:26 +02:00
Javier Martinez
50b3027a24
docs: update docs and capture ( #2029 )
...
* docs: update Readme
* style: refactor image
* docs: change important to tip
2024-08-01 10:01:22 +02:00
Javier Martinez
54659588b5
fix: nomic embeddings ( #2030 )
...
* fix: allow to configure trust_remote_code
based on: https://github.com/zylon-ai/private-gpt/issues/1893#issuecomment-2118629391
* fix: nomic hf embeddings
2024-08-01 09:43:30 +02:00
Javier Martinez
8119842ae6
feat(recipe): add our first recipe Summarize ( #2028 )
...
* feat: add summary recipe
* test: add summary tests
* docs: move all recipes docs
* docs: add recipes and summarize doc
* docs: update openapi reference
* refactor: split method in two method (summary)
* feat: add initial summarize ui
* feat: add mode explanation
* fix: mypy
* feat: allow to configure async property in summarize
* refactor: move modes to enum and update mode explanations
* docs: fix url
* docs: remove list-llm pages
* docs: remove double header
* fix: summary description
2024-07-31 16:53:27 +02:00
Javier Martinez
40638a18a5
fix: unify embedding models ( #2027 )
...
* feat: unify embedding model to nomic
* docs: add embedding dimensions mismatch
* docs: fix fern
2024-07-31 14:35:46 +02:00
Javier Martinez
9027d695c1
feat: make llama3.1 as default ( #2022 )
...
* feat: change ollama default model to llama3.1
* chore: bump versions
* feat: Change default model in local mode to llama3.1
* chore: make sure last poetry version is used
* fix: mypy
* fix: do not add BOS (with last llamacpp-python version)
2024-07-31 14:35:36 +02:00
Javier Martinez
e54a8fe043
fix: prevent to ingest local files (by default) ( #2010 )
...
* feat: prevent to local ingestion (by default) and add white-list
* docs: add local ingestion warning
* docs: add missing comment
* fix: update exception error
* fix: black
2024-07-31 14:33:46 +02:00
Javier Martinez
1020cd5328
fix: light mode ( #2025 )
2024-07-31 12:59:31 +02:00
Quentin McGaw
65c5a1708b
chore(docker): dockerfiles improvements and fixes ( #1792 )
...
* `UID` and `GID` build arguments for `worker` user
* `POETRY_EXTRAS` build argument with default values
* Copy `Makefile` for `make ingest` command
* Do NOT copy markdown files
I doubt anyone reads a markdown file within a Docker image
* Fix PYTHONPATH value
* Set home directory to `/home/worker` when creating user
* Combine `ENV` instructions together
* Define environment variables with their defaults
- For documentation purposes
- Reflect defaults set in settings-docker.yml
* `PGPT_EMBEDDING_MODE` to define embedding mode
* Remove ineffective `python3 -m pipx ensurepath`
* Use `&&` instead of `;` to chain commands to detect failure better
* Add `--no-root` flag to poetry install commands
* Set PGPT_PROFILES to docker
* chore: remove envs
* chore: update to use ollama in docker-compose
* chore: don't copy makefile
* chore: don't copy fern
* fix: tiktoken cache
* fix: docker compose port
* fix: ffmpy dependency (#2020 )
* fix: ffmpy dependency
* fix: block ffmpy to commit sha
* feat(llm): autopull ollama models (#2019 )
* chore: update ollama (llm)
* feat: allow to autopull ollama models
* fix: mypy
* chore: install always ollama client
* refactor: check connection and pull ollama method to utils
* docs: update ollama config with autopulling info
...
* chore: autopull ollama models
* chore: add GID/UID comment
...
---------
Co-authored-by: Javier Martinez <javiermartinezalvarez98@gmail.com>
2024-07-30 17:59:38 +02:00
Robert Hirsch
d080969407
added llama3 prompt ( #1962 )
...
* added llama3 prompt
* more fixes to pass tests; changed type VectorStore -> BasePydanticVectorStore, see https://github.com/run-llama/llama_index/blob/main/CHANGELOG.md#2024-05-14
* fix: new llama3 prompt
---------
Co-authored-by: Javier Martinez <javiermartinezalvarez98@gmail.com>
2024-07-29 17:28:00 +02:00
Javier Martinez
d4375d078f
fix(ui): gradio bug fixes ( #2021 )
...
* fix: when two user messages were sent
* fix: add source divider
* fix: add favicon
* fix: add zylon link
* refactor: update label
2024-07-29 16:48:16 +02:00
Javier Martinez
20bad17c98
feat(llm): autopull ollama models ( #2019 )
...
* chore: update ollama (llm)
* feat: allow to autopull ollama models
* fix: mypy
* chore: install always ollama client
* refactor: check connection and pull ollama method to utils
* docs: update ollama config with autopulling info
2024-07-29 13:25:42 +02:00
Javier Martinez
dabf556dae
fix: ffmpy dependency ( #2020 )
...
* fix: ffmpy dependency
* fix: block ffmpy to commit sha
2024-07-29 11:56:57 +02:00
Iván Martínez
05a986231c
Add proper param to demo urls ( #2007 )
2024-07-22 14:44:03 +02:00
Javier Martinez
b62669784b
docs: update welcome page ( #2004 )
2024-07-18 14:42:39 +02:00
Javier Martinez
2c78bb2958
docs: add PR and issue templates ( #2002 )
...
* chore: add pull request template
* chore: add issue templates
* chore: require more information in bugs
2024-07-18 12:56:10 +02:00
Iván Martínez
90d211c5cd
Update README.md ( #2003 )
...
* Update README.md
Remove the outdated contact form and point to Zylon website for those looking for a ready-to-use enterprise solution built on top of PrivateGPT
* Update README.md
Update text to address the comments
* Update README.md
Improve text
2024-07-18 12:11:24 +02:00
Jackson
43cc31f740
feat(vectordb): Milvus vector db Integration ( #1996 )
...
* integrate Milvus into Private GPT
* adjust milvus settings
* update doc info and reformat
* adjust milvus initialization
* adjust import error
* mionr update
* adjust format
* adjust the db storing path
* update doc
2024-07-18 10:55:45 +02:00
Javier Martinez
4523a30c8f
feat(docs): update documentation and fix preview-docs ( #2000 )
...
* docs: add missing configurations
* docs: change HF embeddings by ollama
* docs: add disclaimer about Gradio UI
* docs: improve readability in concepts
* docs: reorder `Fully Local Setups`
* docs: improve setup instructions
* docs: prevent have duplicate documentation and use table to show different options
* docs: rename privateGpt to PrivateGPT
* docs: update ui image
* docs: remove useless header
* docs: convert to alerts ingestion disclaimers
* docs: add UI alternatives
* docs: reference UI alternatives in disclaimers
* docs: fix table
* chore: update doc preview version
* chore: add permissions
* chore: remove useless line
* docs: fixes
...
2024-07-18 10:06:51 +02:00
Javier Martinez
01b7ccd064
fix(config): make tokenizer optional and include a troubleshooting doc ( #1998 )
...
* docs: add troubleshooting
* fix: pass HF token to setup script and prevent to download tokenizer when it is empty
* fix: improve log and disable specific tokenizer by default
* chore: change HF_TOKEN environment to be aligned with default config
* ifx: mypy
2024-07-17 10:06:27 +02:00
Javier Martinez
15f73dbc48
docs: update repo links, citations ( #1990 )
...
* docs: update project links
...
* docs: update citation
2024-07-09 10:03:57 +02:00
fern
187bc9320e
(feat): add github button ( #1989 )
...
Co-authored-by: chdeskur <chdeskur@gmail.com>
2024-07-09 08:48:47 +02:00
Marco Braga
dde02245bc
fix(docs): Fix concepts.mdx referencing to installation page ( #1779 )
...
* Fix/update concepts.mdx referencing to installation page
The link for `/installation` is broken in the "Main Concepts" page.
The correct path would be `./installation` or maybe `/installation/getting-started/installation`
* fix: docs
---------
Co-authored-by: Javier Martinez <javiermartinezalvarez98@gmail.com>
2024-07-08 16:19:50 +02:00
Mart
067a5f144c
feat(docs): Fix setup docu ( #1926 )
...
* Update settings.mdx
* docs: add cmd
---------
Co-authored-by: Javier Martinez <javiermartinezalvarez98@gmail.com>
2024-07-08 16:19:16 +02:00
Proger666
2612928839
feat(vectorstore): Add clickhouse support as vectore store ( #1883 )
...
* Added ClickHouse vector sotre support
* port fix
* updated lock file
* fix: mypy
* fix: mypy
---------
Co-authored-by: Valery Denisov <valerydenisov@double.cloud>
Co-authored-by: Javier Martinez <javiermartinezalvarez98@gmail.com>
2024-07-08 16:18:22 +02:00
