End of training
Browse files- config.json +2 -3
- generation_config.json +1 -1
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -20,17 +20,16 @@
|
|
| 20 |
"num_key_value_heads": 8,
|
| 21 |
"pretraining_tp": 1,
|
| 22 |
"rms_norm_eps": 1e-05,
|
| 23 |
-
"
|
| 24 |
"factor": 32.0,
|
| 25 |
"high_freq_factor": 4.0,
|
| 26 |
"low_freq_factor": 1.0,
|
| 27 |
"original_max_position_embeddings": 8192,
|
| 28 |
-
"rope_theta": 500000.0,
|
| 29 |
"rope_type": "llama3"
|
| 30 |
},
|
| 31 |
"rope_theta": 500000.0,
|
| 32 |
"tie_word_embeddings": true,
|
| 33 |
-
"transformers_version": "
|
| 34 |
"use_cache": true,
|
| 35 |
"vocab_size": 128256
|
| 36 |
}
|
|
|
|
| 20 |
"num_key_value_heads": 8,
|
| 21 |
"pretraining_tp": 1,
|
| 22 |
"rms_norm_eps": 1e-05,
|
| 23 |
+
"rope_scaling": {
|
| 24 |
"factor": 32.0,
|
| 25 |
"high_freq_factor": 4.0,
|
| 26 |
"low_freq_factor": 1.0,
|
| 27 |
"original_max_position_embeddings": 8192,
|
|
|
|
| 28 |
"rope_type": "llama3"
|
| 29 |
},
|
| 30 |
"rope_theta": 500000.0,
|
| 31 |
"tie_word_embeddings": true,
|
| 32 |
+
"transformers_version": "4.57.1",
|
| 33 |
"use_cache": true,
|
| 34 |
"vocab_size": 128256
|
| 35 |
}
|
generation_config.json
CHANGED
|
@@ -5,5 +5,5 @@
|
|
| 5 |
"eos_token_id": 128001,
|
| 6 |
"temperature": 0.6,
|
| 7 |
"top_p": 0.9,
|
| 8 |
-
"transformers_version": "
|
| 9 |
}
|
|
|
|
| 5 |
"eos_token_id": 128001,
|
| 6 |
"temperature": 0.6,
|
| 7 |
"top_p": 0.9,
|
| 8 |
+
"transformers_version": "4.57.1"
|
| 9 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4943274328
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0418036d8baaf75567b7d4f31a4172e1acca7f30f223c273871359577a5f9cab
|
| 3 |
size 4943274328
|