medmekk
/

Llama-3.2-1B-FBGEMM-FP8

Model card Files Files and versions Community

medmekk HF staff commited on 7 days ago

Commit

a968a68

·

verified ·

1 Parent(s): 95e4cda

Update config.json

Files changed (1) hide show

config.json +1 -1

config.json CHANGED Viewed

@@ -22,7 +22,7 @@
   "quantization_config": {
     "activation_scale_ub": 1200.0,
     "modules_to_not_convert": null,
-    "quant_method": "fbgemm_fp8"
   },
   "rms_norm_eps": 1e-05,
   "rope_scaling": {

   "quantization_config": {
     "activation_scale_ub": 1200.0,
     "modules_to_not_convert": null,
+    "quant_method": "fbgemm_example"
   },
   "rms_norm_eps": 1e-05,
   "rope_scaling": {