Fetching metadata from the HF Docker repository...
-
Refined the Links
-
Removed older Mistral references
-
Added Blog Post link
-
Added LinkedIn Post link
-
Added the YouTube Demo Link
-
Further bug fix
-
Fixed Syntax error
-
Removed Keyword Coverage from PDF report as well
-
Removed Coverage Percentage inside feedback
-
Removed Score Badge
-
Removed Keyword Coverage Percentage
-
About Page update
-
Tuned build_prompt further
-
Back to 7B
-
Trying Qwen/Qwen2.5-14B-Instruct model to test performance
-
Update scorer.py to be smarter in scoring
-
Delete white_theme_agenda.md
-
Delete Project_Agenda.md
-
Delete Pointers to add to Interview Coach.txt
-
Delete Project_Agenda.pdf
-
Delete conversation_summary_phase-1.md
-
Delete Project_Safety_Measures.md
-
Delete Project_Safety_Measures.txt
-
Corrected tags
-
Updated README to have appropriate tags and thumbnail also added
-
Deleted conversation_summary.md
-
Updated UI to say "Powered by Qwen2.5-7B-Instruct"
-
fixed minor comment error, not changed code
-
Updated validator.py to be Qwn compatible
Since Qwen doesnt use mistral's token tags(like <s>[INST]), we remove them to ensure clean text gen -
Updated engine.py to make message payload OpenAI-compatible for Qwen2.5B
-
Updated config.py to call Qwen2.5-7B-Instruct
-
Changed ask_llm() so it uses client.chat_completion() instead of text_generation, as the models were saying text gen not supprted
-
Corrected Indentation of _gatekeeper()
-
Updates _gatekeepr() to be Mistral 7B v02 compatible
Mistral needs prompt to be wrapped inside <s> and [INST] tags hence the change -
Moved the css and theme call out of interview_coach INTO app.py
-
moved custom css and theme call out of gr.Blocks()
-
Refining ask_llm function to help in debugging
-
Switched to Mistral 7B v02 (text generation model)
-
Making client Qwen compatible
-
Updated _generate to be Qwen compatible
-
Update config.py
-
Update engine.py
-
Update engine.py
-
Update engine.py
-
Update engine.py
-
Update engine.py
-
Update engine.py
-
Skip Ollama check on Hugging Face Spaces
ishan-25
-
Propagate API exceptions and show connection errors in Gradio UI
-
Fix HF Client Validation, venv package mismatch, Ollama timeouts
