Downloads · 30 days
4
6% of all-time downloads
friendliai-q1
drranny/friendliai-q1
friendliai-q1 is a text generation model from drranny. Use it when you need the model to write or continue text. It is set up for transformers.
This model could not be used to run a functional /chat/completions API server.
All-time downloads
65
Public
Repo size
11.4 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.json14.2 MB · 89%
At a glance
- Task
- Text Generation
- Library
- transformers
- Model type
- qwen3
- Access
- Public
- Created
- May 9, 2026
- Updated
- May 9, 2026
- SHA
- 190fadad
Try a prompt
Base models
Model card
From the Hugging Face model README
Bug Fix Report
Problem
This model could not be used to run a functional /chat/completions API server.
Root Cause
The root cause was a missing chat_template entry in tokenizer_config.json.
Without a valid chat template:
- OpenAI-style
/chat/completionsformatting could not be constructed correctly. - Conversational role formatting (system/user/assistant turns) failed.
- Qwen3 thinking-mode activation logic (
enable_thinking) was never triggered.
Fix
No model weights or architecture parameters were modified. Only the tokenizer-side conversational formatting configuration was corrected.
| File | Field | Before | After |
|---|---|---|---|
tokenizer_config.json | chat_template | missing | Restored Qwen3 Jinja2 template |
The minimal fix was to restore the correct Qwen3 chat template from the official
Qwen/Qwen3-8B tokenizer configuration.
Connection to reasoning_effort
This issue also explains why reasoning_effort appeared ineffective:
the missing template prevented Qwen3 thinking-mode control logic
(enable_thinking) from activating correctly.
- Task
- Text Generation
- Library
- transformers
- Type
- qwen3
- Created
- May 9, 2026
- Updated
- May 9, 2026