28 lines
754 B
JSON
28 lines
754 B
JSON
{
|
|
"_name_or_path": "/weka/oe-adapt-default/costah/models/olmo2/1210_13b_gsm_beta_0.03_lr_3e-7_7470_checkpoints/step_180/",
|
|
"architectures": [
|
|
"Olmo2ForCausalLM"
|
|
],
|
|
"attention_bias": false,
|
|
"attention_dropout": 0.0,
|
|
"eos_token_id": 100257,
|
|
"hidden_act": "silu",
|
|
"hidden_size": 5120,
|
|
"initializer_range": 0.02,
|
|
"intermediate_size": 13824,
|
|
"max_position_embeddings": 4096,
|
|
"model_type": "olmo2",
|
|
"num_attention_heads": 40,
|
|
"num_hidden_layers": 40,
|
|
"num_key_value_heads": 40,
|
|
"pad_token_id": 100277,
|
|
"rms_norm_eps": 1e-06,
|
|
"rope_scaling": null,
|
|
"rope_theta": 500000,
|
|
"tie_word_embeddings": false,
|
|
"torch_dtype": "bfloat16",
|
|
"transformers_version": "4.47.0.dev0",
|
|
"use_cache": false,
|
|
"vocab_size": 100352
|
|
}
|