[editor] HuggingFaceTextGenerationParserPromptSchema #755
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# [editor] PaLMChatParserPromptSchema Implement the Prompt Schema for PaLM Chat, using the supported params with defaults from https://cloud.google.com/vertex-ai/docs/generative-ai/model-reference/text-chat, selecting those that are defined in the `refine_chat_completion_params` of palm.py: ``` def refine_chat_completion_params(model_settings): # completion parameters to be used for Palm's chat completion api # messages handled seperately supported_keys = { "candidate_count", "examples", "model", "temperature", "top_k", "top_p", "context", } ``` <img width="1853" alt="Screenshot 2024-01-04 at 1 59 24 PM" src="https://github.com/lastmile-ai/aiconfig/assets/5060851/50e08b43-2956-4fa0-98e4-4ab165efb6a9">
rholinshead
requested review from
saqadri,
suyoglastmileai,
Ankush-lastmile,
jonathanlastmileai and
rossdanlm
as code owners
# [editor] HuggingFaceTextGenerationParserPromptSchema Adding the PromptSchema for HuggingFaceTextGenerationParser, with supported properties from `refine_chat_completion_params` in the parser implementation. Types obtained from https://github.com/huggingface/huggingface_hub/blob/a331e82aad1bc63038194611236db28fa013814c/src/huggingface_hub/inference/_client.py#L1206 and defaults obtained from https://huggingface.co/docs/api-inference/detailed_parameters if they're listed <img width="1080" alt="Screenshot 2024-01-04 at 3 19 05 PM" src="https://github.com/lastmile-ai/aiconfig/assets/5060851/8cb40459-1f7c-42b2-8b35-0333c380be01"> Note, this is for the default/core parser which uses inference API. We'll need to add the other prompt schemas for the extension models for gradio. Will update the hf prompt UX to support actual 'model' in a subsequent diff
|
Landing to unblock bug bashing, will fix forward as needed |
rholinshead
added a commit
that referenced
this pull request
Jan 4, 2024
# [editor] Fix Model Search to Ignore Casing Previously, the model search was doing exact match, including capitalization. So, searching 'hu' would not show the "HuggingFaceTextGenerationParser". Fix by making both search and model names lowercase when filtering: <img width="354" alt="Screenshot 2024-01-04 at 3 35 46 PM" src="https://github.com/lastmile-ai/aiconfig/assets/5060851/10baf717-6192-4de8-bab6-046f678ac9ef"> --- Stack created with [Sapling](https://sapling-scm.com). Best reviewed with [ReviewStack](https://reviewstack.dev/lastmile-ai/aiconfig/pull/756). * __->__ #756 * #755 * #754
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
[editor] HuggingFaceTextGenerationParserPromptSchema
[editor] HuggingFaceTextGenerationParserPromptSchema
Adding the PromptSchema for HuggingFaceTextGenerationParser, with supported properties from
refine_chat_completion_paramsin the parser implementation. Types obtained from https://github.com/huggingface/huggingface_hub/blob/a331e82aad1bc63038194611236db28fa013814c/src/huggingface_hub/inference/_client.py#L1206 and defaults obtained from https://huggingface.co/docs/api-inference/detailed_parameters if they're listedNote, this is for the default/core parser which uses inference API. We'll need to add the other prompt schemas for the extension models for gradio.
Will update the hf prompt UX to support actual 'model' in a subsequent diff
Stack created with Sapling. Best reviewed with ReviewStack.