Commit
•
a878ffa
1
Parent(s):
1f64ce2
Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
---
|
| 3 |
+
|
| 4 |
+
license: apache-2.0
|
| 5 |
+
datasets:
|
| 6 |
+
- arcee-ai/EvolKit-20k
|
| 7 |
+
base_model:
|
| 8 |
+
- Qwen/Qwen2.5-1.5B
|
| 9 |
+
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
[](https://hf.co/QuantFactory)
|
| 13 |
+
|
| 14 |
+
|
| 15 |
+
# QuantFactory/EVA-D-Qwen2.5-1.5B-v0.0-GGUF
|
| 16 |
+
This is quantized version of [EVA-UNIT-01/EVA-D-Qwen2.5-1.5B-v0.0](https://huggingface.co/EVA-UNIT-01/EVA-D-Qwen2.5-1.5B-v0.0) created using llama.cpp
|
| 17 |
+
|
| 18 |
+
# Original Model Card
|
| 19 |
+
|
| 20 |
+
# EVA-D Qwen2.5-1.5B v0.0
|
| 21 |
+
|
| 22 |
+
<p>
|
| 23 |
+
An experimental online logit distillation of EVA-Qwen2.5-14B-v0.1 into Qwen2.5-1.5B. Should work as a RP/storywriting specialist, but don't expect superb performance from it, due to it's small size. All in all, it was a fun experiment to do.<br>
|
| 24 |
+
</p>
|
| 25 |
+
|
| 26 |
+
<p>Note: using quantized KV cache with Qwen2.5 <b>is not recommended</b> and can lead to degraded output quality. On the other hand, Qwen's KV cache is already light enough, so using f16 for it shouldn't be problematic.</p>
|
| 27 |
+
|
| 28 |
+
<p>
|
| 29 |
+
<p>Prompt format is ChatML.</p><br>
|
| 30 |
+
<h3>Recommended sampler values:</h3>
|
| 31 |
+
<ul>
|
| 32 |
+
<li>Temperature: 1</li>
|
| 33 |
+
<li>Min-P: 0.02</li>
|
| 34 |
+
</ul>
|
| 35 |
+
|
| 36 |
+
<h3>Recommended SillyTavern presets (via CalamitousFelicitousness):</h3>
|
| 37 |
+
|
| 38 |
+
- [Context](https://huggingface.co/EVA-UNIT-01/EVA-Yi-1.5-9B-32K-V1/blob/main/%5BChatML%5D%20Roleplay-v1.9%20Context.json)
|
| 39 |
+
- [Instruct and System Prompt](https://huggingface.co/EVA-UNIT-01/EVA-Yi-1.5-9B-32K-V1/blob/main/%5BChatML%5D%20Roleplay-v1.9%20Instruct.json)
|
| 40 |
+
</p>
|
| 41 |
+
|
| 42 |
+
<p>
|
| 43 |
+
<br>
|
| 44 |
+
<h3>
|
| 45 |
+
Distillation data:
|
| 46 |
+
</h3>
|
| 47 |
+
<ul>
|
| 48 |
+
<li>Arcee.AI's <a href=https://huggingface.co/datasets/arcee-ai/EvolKit-20k>EvolKit-20k</a> dataset, which is specifically made for knowledge distillation purposes.</li>
|
| 49 |
+
</ul>
|
| 50 |
+
<h3>
|
| 51 |
+
Training time and hardware:
|
| 52 |
+
</h3>
|
| 53 |
+
<ul><li>1.8 hours on 8xA100 SXM, provided by Garg</li></ul><br>
|
| 54 |
+
</p>
|
| 55 |
+
<p>Model was trained by Kearm and Auri.</p>
|
| 56 |
+
<h4>Special thanks:</h4><ul>
|
| 57 |
+
<li><b>to Garg for generously providing 8xA100 SXM node for this experiment!</b></li>
|
| 58 |
+
<li>to Arcee.AI for creating DistillKit and EvolKit-20k dataset, which were used to create this model.</li>
|
| 59 |
+
<li>and to Allura-org for support and feedback on EVA models.</li></ul>
|