trl-internal-testing
/

tiny-CohereForCausalLM

Text Generation

text-generation-inference

Model card Files Files and versions

qgallouedec HF Staff commited on Sep 24

Commit

ee508c4

·

verified ·

1 Parent(s): a533b44

Upload CohereForCausalLM

Files changed (2) hide show

config.json +4 -4
model.safetensors +2 -2

config.json CHANGED Viewed

@@ -5,14 +5,14 @@
   "attention_bias": false,
   "attention_dropout": 0.0,
   "bos_token_id": 5,
-  "dtype": "bfloat16",
   "eos_token_id": 255001,
   "hidden_act": "silu",
   "hidden_size": 8,
   "initializer_range": 0.02,
   "intermediate_size": 32,
   "layer_norm_eps": 1e-05,
-  "logit_scale": 0.125,
   "max_position_embeddings": 8192,
   "model_type": "cohere",
   "num_attention_heads": 4,
@@ -20,9 +20,9 @@
   "num_key_value_heads": 2,
   "pad_token_id": 0,
   "rope_scaling": null,
-  "rope_theta": 10000,
   "transformers_version": "4.57.0.dev0",
   "use_cache": true,
   "use_qk_norm": false,
-  "vocab_size": 256000
 }

   "attention_bias": false,
   "attention_dropout": 0.0,
   "bos_token_id": 5,
+  "dtype": "float32",
   "eos_token_id": 255001,
   "hidden_act": "silu",
   "hidden_size": 8,
   "initializer_range": 0.02,
   "intermediate_size": 32,
   "layer_norm_eps": 1e-05,
+  "logit_scale": 0.0625,
   "max_position_embeddings": 8192,
   "model_type": "cohere",
   "num_attention_heads": 4,
   "num_key_value_heads": 2,
   "pad_token_id": 0,
   "rope_scaling": null,
+  "rope_theta": 10000.0,
   "transformers_version": "4.57.0.dev0",
   "use_cache": true,
   "use_qk_norm": false,
+  "vocab_size": 255029
 }

model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6bf912887f8d5baeed3097feb2c594f3dd2f247623e80ae4b089a150ad32490c
-size 4101760

 version https://git-lfs.github.com/spec/v1
+oid sha256:29b5963e75fcb61c9eb02696dfbc183e5adea11da21b9cc1eb394d5b16faf3ee
+size 8170560