Views
No views yet
| Parameter | Value |
|---|---|
| start_layer_index | 34 |
| end_layer_index | 73 |
| preserve_good_behavior_weight | 0.5584 |
| steer_bad_behavior_weight | 0.0005 |
| overcorrect_relative_weight | 0.9231 |
| neighbor_count | 8 |
| Metric | This model | Original model (Zyphra/ZAYA1-74B-preview) |
|---|---|---|
| KL divergence | 0.0095 | 0 (by definition) |
| Refusals | 0/100 | 96/100 |
zaya1-pr branch from our fork of vllm library (the command will trigger a full build of vLLM from source):pip install "vllm @ git+https://github.com/Zyphra/vllm.git@zaya1-pr"zaya1 branch from our fork of transformers library as well:pip install "transformers @ git+https://github.com/Zyphra/transformers.git@zaya1"1vllm serve Zyphra/ZAYA1-74B-Preview --port 8010 \
2 --mamba-cache-dtype float32 --dtype bfloat16 \
3 --reasoning-parser qwen3 --enable-auto-tool-choice --tool-call-parser zaya_xml-dp 8 -ep to run with DP=EP=8.curl like in the following example:1curl http://localhost:8010/v1/chat/completions \
2 -H "Content-Type: application/json" \
3 -d '{
4 "model": "Zyphra/ZAYA1-74B-Preview",
5 "messages": [
6 {"role": "system", "content": "You are a helpful assistant."},
7 {"role": "user", "content": "Hello. How is it going?"}
8 ]
9 }'