Checkpoint step 2000, loss=5.3471
Browse files- config.json +2 -2
- generation_config.json +1 -1
- model.safetensors +2 -2
config.json
CHANGED
|
@@ -6,6 +6,7 @@
|
|
| 6 |
"attention_dropout": 0.0,
|
| 7 |
"attention_multiplier": 0.015625,
|
| 8 |
"bos_token_id": 100257,
|
|
|
|
| 9 |
"embedding_multiplier": 12,
|
| 10 |
"eos_token_id": 100257,
|
| 11 |
"hidden_act": "silu",
|
|
@@ -71,8 +72,7 @@
|
|
| 71 |
"router_aux_loss_coef": 0.01,
|
| 72 |
"shared_intermediate_size": 2048,
|
| 73 |
"tie_word_embeddings": true,
|
| 74 |
-
"
|
| 75 |
-
"transformers_version": "4.53.3",
|
| 76 |
"use_cache": false,
|
| 77 |
"vocab_size": 116481
|
| 78 |
}
|
|
|
|
| 6 |
"attention_dropout": 0.0,
|
| 7 |
"attention_multiplier": 0.015625,
|
| 8 |
"bos_token_id": 100257,
|
| 9 |
+
"dtype": "bfloat16",
|
| 10 |
"embedding_multiplier": 12,
|
| 11 |
"eos_token_id": 100257,
|
| 12 |
"hidden_act": "silu",
|
|
|
|
| 72 |
"router_aux_loss_coef": 0.01,
|
| 73 |
"shared_intermediate_size": 2048,
|
| 74 |
"tie_word_embeddings": true,
|
| 75 |
+
"transformers_version": "4.57.3",
|
|
|
|
| 76 |
"use_cache": false,
|
| 77 |
"vocab_size": 116481
|
| 78 |
}
|
generation_config.json
CHANGED
|
@@ -3,5 +3,5 @@
|
|
| 3 |
"bos_token_id": 100257,
|
| 4 |
"eos_token_id": 100257,
|
| 5 |
"pad_token_id": 100256,
|
| 6 |
-
"transformers_version": "4.
|
| 7 |
}
|
|
|
|
| 3 |
"bos_token_id": 100257,
|
| 4 |
"eos_token_id": 100257,
|
| 5 |
"pad_token_id": 100256,
|
| 6 |
+
"transformers_version": "4.57.3"
|
| 7 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:524bc93bc87da1cb98290fdd46bc975479b705726b8d837151be5007a06850ed
|
| 3 |
+
size 737818416
|