Upload quant-forge artifacts
Browse files
README.md
CHANGED
|
@@ -1,3 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
# Neooooo/qf-integration-test
|
| 2 |
|
| 3 |
## QuantForge Metadata
|
|
@@ -11,6 +24,9 @@
|
|
| 11 |
|
| 12 |
## Accuracy (BF16 vs NVFP4)
|
| 13 |
|
|
|
|
|
|
|
|
|
|
| 14 |
| Task | Metric | BF16 | NVFP4 | Recovery |
|
| 15 |
|---|---:|---:|---:|---:|
|
| 16 |
| arc_challenge | acc,none | 0.5000 | n/a | n/a |
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: Qwen/Qwen3-30B-A3B
|
| 3 |
+
datasets:
|
| 4 |
+
- HuggingFaceH4/ultrachat_200k
|
| 5 |
+
library_name: transformers
|
| 6 |
+
pipeline_tag: text-generation
|
| 7 |
+
quantized_by: QuantForge
|
| 8 |
+
tags:
|
| 9 |
+
- quantforge
|
| 10 |
+
- quantized
|
| 11 |
+
- nvfp4
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
# Neooooo/qf-integration-test
|
| 15 |
|
| 16 |
## QuantForge Metadata
|
|
|
|
| 24 |
|
| 25 |
## Accuracy (BF16 vs NVFP4)
|
| 26 |
|
| 27 |
+
_Recovery status: partial_
|
| 28 |
+
_Details: Quantized evaluation unavailable due to current vLLM MoE quantization compatibility in this environment._
|
| 29 |
+
|
| 30 |
| Task | Metric | BF16 | NVFP4 | Recovery |
|
| 31 |
|---|---:|---:|---:|---:|
|
| 32 |
| arc_challenge | acc,none | 0.5000 | n/a | n/a |
|