Upload hf_quant_config.json with huggingface_hub
Browse files- hf_quant_config.json +10 -0
hf_quant_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"quantization": {
|
| 3 |
+
"exclude_modules": [
|
| 4 |
+
"lm_head",
|
| 5 |
+
"model.embed_tokens"
|
| 6 |
+
],
|
| 7 |
+
"kv_cache_quant_algo": null,
|
| 8 |
+
"quant_algo": "FP8"
|
| 9 |
+
}
|
| 10 |
+
}
|