Commit Graph

30 Commits

Author SHA1 Message Date
Timothy Jaeryang Baek
eeb00a5ca2 chore: format 2025-02-20 01:01:29 -08:00
Timothy Jaeryang Baek
13e58c1b64
Merge pull request #10360 from ferret99gt/ollama-option-conversion-fixes
fix: Update Ollama option handling in payload.py's convert_payload_openai_to_ollama
2025-02-19 22:44:07 -08:00
Timothy Jaeryang Baek
efefff71d1
Merge pull request #10363 from ferret99gt/ollama-model-option-conversion-updates
feat: Update Payload's apply_model_params_to_body_ollama with full Ollama supported options
2025-02-19 12:32:46 -08:00
Simone
6f698073bb Add support for tool message in openai to Ollama message conversion 2025-02-19 21:11:39 +01:00
ferret99gt
5701d6d333 Change the opt dictionary to a mappings dictionary with appropriate casts
This is to bring consistency with apply_model_params_to_body_openai. Both now use a mapping dictionary then call and return apply_model_params_to_body directly.
2025-02-19 09:30:16 -05:00
ferret99gt
fa885c3346 Update remapping logic
We copy the params from from the original key to the new key, then delete it. This is to ensure Ollama only gets valid options.

(Add a comment as well)
2025-02-19 09:25:47 -05:00
ferret99gt
a4249a6351 Remove remapping of frequency_penalty to repeat_penalty
1) Ollama natively supports frequency_penalty, so this is not necessary.
2) Repeat_penalty is being added to Open WebUI in PR #10016, allowing Ollama users to pick which penalty methods they want.
2025-02-19 09:24:38 -05:00
ferret99gt
877d21a029 Move parameter remapping above the opt dictionary.
This is so that any remappings are handled before options are handled by apply_model_params_to_body
2025-02-19 09:23:33 -05:00
ferret99gt
6b2ba73701 Remove tfs_z
See https://github.com/ollama/ollama/blob/main/docs/api.md#request-8

tfs_z was removed from Llama.cpp and Ollama in January 2025.
2025-02-19 09:21:36 -05:00
ferret99gt
cee08e316d Update Ollama request option list with full supported options
See https://github.com/ollama/ollama/blob/main/docs/api.md#request-8

This adds the full suite of supported Ollama options.
2025-02-19 09:21:05 -05:00
ferret99gt
8125b0499b Remove empty ollama_options
1) The empty dictionary from line 176 is never used.
2) Lines 193-194 are not necessary as they were already done at line 177
2025-02-19 08:57:05 -05:00
ferret99gt
57b01cf8fb Fix for system prompt setting
1) Ollama supports sending the system prompt as a parameter, not as an option. (See https://github.com/ollama/ollama/blob/main/docs/api.md#request-8) However, it is in the options dictionary and needs moved to the payload dictionary.
2) After moving the system parameter from ollama_options to ollama_payload, delete it from ollama_options. This is to prevent Ollama throwing a warning about invalid options.
2025-02-19 08:55:11 -05:00
ferret99gt
fea169a9c0 Core fix for num_predict not working.
1) max_tokens was being looked for in openai_payload, but is present in openai_payload['options'], so is never found.
2) After copying the value for max_tokens to num_predict, delete max_tokens from the dictionary. This is to prevent Ollama throwing a warning about invalid option (max_tokens)
2025-02-19 08:52:34 -05:00
ferret99gt
adde37394b Remove parameters that map directly, as they are part of options
1) This may be legacy code?
2) All three of these parameters, temperature, top_p and seed, are found in openai_payload["options"], not openai_payload. They do not need remapped any longer.
2025-02-19 08:44:59 -05:00
ferret99gt
aea8977d05 Remove mapping of max_completion_tokens
1) max_completion_tokens is being looked for in openai_payload, but would be located in openai_payload['options'], so is never found. (This applies to the prior two commits as well).
2) max_completion_tokens is not sent from the frontend, only max_tokens. It does not appear in AdvancedParams.svelte.
2b) Openai.py does use max_completion_tokens, but for o1,o3 models and converts it from max_tokens.
2025-02-19 08:39:33 -05:00
ferret99gt
e6919c3242 Remove mapping of frequency_penalty to repeat_penalty
1) Ollama natively supports frequency_penalty.
2) repeat_penaltywas added to Open Webui in PR #10016 and is not merged to main yet at this time. Once both changes go live, Ollama users can freely choose between frequency/presence penalty, or repeat penalty, as they choose.
2025-02-19 08:33:29 -05:00
ferret99gt
a560f789e4 Remove mapping of presence_penalty to new_topix_penalty
1) Ollama natively supports presence_penalty.
2) new_topic_penalty is not a valid option in Ollama. (See https://github.com/ollama/ollama/blob/main/docs/api.md#request-8)
3) Presence_penalty was added to Open Webui in PR #10016 and is not merged to main yet at this time.
2025-02-19 08:31:57 -05:00
Timothy Jaeryang Baek
63a5ee3adb refac 2025-02-10 11:25:38 -08:00
Timothy Jaeryang Baek
7ba717d171 refac: system prompt template 2025-02-05 14:35:18 -08:00
Timothy Jaeryang Baek
ff84c120f5 fix: api usage 2025-02-05 13:18:17 -08:00
Timothy Jaeryang Baek
8d520eeecd fix: ollama native tool calling 2025-02-04 21:08:26 -08:00
Timothy Jaeryang Baek
cc99673906 refac: prompt variables
Some checks are pending
Deploy to HuggingFace Spaces / check-secret (push) Waiting to run
Deploy to HuggingFace Spaces / deploy (push) Blocked by required conditions
Create and publish Docker images with specific build args / build-main-image (linux/amd64) (push) Waiting to run
Create and publish Docker images with specific build args / build-main-image (linux/arm64) (push) Waiting to run
Create and publish Docker images with specific build args / build-cuda-image (linux/amd64) (push) Waiting to run
Create and publish Docker images with specific build args / build-cuda-image (linux/arm64) (push) Waiting to run
Create and publish Docker images with specific build args / build-ollama-image (linux/amd64) (push) Waiting to run
Create and publish Docker images with specific build args / build-ollama-image (linux/arm64) (push) Waiting to run
Create and publish Docker images with specific build args / merge-main-images (push) Blocked by required conditions
Create and publish Docker images with specific build args / merge-cuda-images (push) Blocked by required conditions
Create and publish Docker images with specific build args / merge-ollama-images (push) Blocked by required conditions
Python CI / Format Backend (3.11) (push) Waiting to run
Frontend Build / Format & Build Frontend (push) Waiting to run
Frontend Build / Frontend Unit Tests (push) Waiting to run
2025-01-29 21:56:51 -08:00
Timothy Jaeryang Baek
aaf97b85dc enh: reasoning_effort param support for openai 2025-01-22 12:07:04 -08:00
Timothy Jaeryang Baek
79d7e67e51 fix: ollama options 2024-12-29 16:30:36 -08:00
envision3d
16d900247a feat: pass structured output format through to ollama 2024-12-17 09:33:32 -06:00
Timothy J. Baek
9f285fb2fb feat: arena models 2024-10-22 03:16:48 -07:00
Timothy J. Baek
b4e73c7f19 refac: convert_payload_openai_to_ollama 2024-10-21 04:10:28 -07:00
Timothy J. Baek
41926172d3 fix/refac: use ollama /api/chat endpoint for tasks 2024-09-21 00:30:13 +02:00
Michał Surmaczewski
8e6ea49e0e fix: incorrect casting of top_p and frequency_penalty 2024-09-08 17:52:58 +02:00
Timothy J. Baek
ff46fe2b4a refac 2024-09-07 03:09:57 +01:00