CohereForAI
/

c4ai-command-r-v01-4bit

Text Generation

Inference Endpoints

text-generation-inference

4-bit precision

Model card Files Files and versions Community

ahmetustun commited on Mar 18

Commit

0e797c8

•

1 Parent(s): 43591b7

Update config.json

Files changed (1) hide show

config.json +2 -2

config.json CHANGED Viewed

@@ -28,7 +28,7 @@
   "quantization_config": {
     "_load_in_4bit": true,
     "_load_in_8bit": false,
-    "bnb_4bit_compute_dtype": "float32",
     "bnb_4bit_quant_storage": "uint8",
     "bnb_4bit_quant_type": "fp4",
     "bnb_4bit_use_double_quant": false,
@@ -41,7 +41,7 @@
     "quant_method": "bitsandbytes"
   },
   "rope_theta": 8000000.0,
-  "torch_dtype": "bfloat16",
   "transformers_version": "4.39.0.dev0",
   "use_cache": true,
   "vocab_size": 256000

   "quantization_config": {
     "_load_in_4bit": true,
     "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "float16",
     "bnb_4bit_quant_storage": "uint8",
     "bnb_4bit_quant_type": "fp4",
     "bnb_4bit_use_double_quant": false,
     "quant_method": "bitsandbytes"
   },
   "rope_theta": 8000000.0,
+  "torch_dtype": "float16",
   "transformers_version": "4.39.0.dev0",
   "use_cache": true,
   "vocab_size": 256000