Delete config.json with huggingface_hub
Browse files- config.json +0 -80
config.json
DELETED
|
@@ -1,80 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"_name_or_path": "TheDrummer/Anubis-70B-v1",
|
| 3 |
-
"architectures": [
|
| 4 |
-
"LlamaForCausalLM"
|
| 5 |
-
],
|
| 6 |
-
"attention_bias": false,
|
| 7 |
-
"attention_dropout": 0.0,
|
| 8 |
-
"bos_token_id": 128000,
|
| 9 |
-
"eos_token_id": 128009,
|
| 10 |
-
"head_dim": 128,
|
| 11 |
-
"hidden_act": "silu",
|
| 12 |
-
"hidden_size": 8192,
|
| 13 |
-
"initializer_range": 0.02,
|
| 14 |
-
"intermediate_size": 28672,
|
| 15 |
-
"max_position_embeddings": 131072,
|
| 16 |
-
"mlp_bias": false,
|
| 17 |
-
"model_type": "llama",
|
| 18 |
-
"num_attention_heads": 64,
|
| 19 |
-
"num_hidden_layers": 80,
|
| 20 |
-
"num_key_value_heads": 8,
|
| 21 |
-
"pad_token_id": 128004,
|
| 22 |
-
"pretraining_tp": 1,
|
| 23 |
-
"quantization_config": {
|
| 24 |
-
"config_groups": {
|
| 25 |
-
"group_0": {
|
| 26 |
-
"input_activations": {
|
| 27 |
-
"actorder": null,
|
| 28 |
-
"block_structure": null,
|
| 29 |
-
"dynamic": true,
|
| 30 |
-
"group_size": null,
|
| 31 |
-
"num_bits": 8,
|
| 32 |
-
"observer": null,
|
| 33 |
-
"observer_kwargs": {},
|
| 34 |
-
"strategy": "token",
|
| 35 |
-
"symmetric": true,
|
| 36 |
-
"type": "float"
|
| 37 |
-
},
|
| 38 |
-
"output_activations": null,
|
| 39 |
-
"targets": [
|
| 40 |
-
"Linear"
|
| 41 |
-
],
|
| 42 |
-
"weights": {
|
| 43 |
-
"actorder": null,
|
| 44 |
-
"block_structure": null,
|
| 45 |
-
"dynamic": false,
|
| 46 |
-
"group_size": null,
|
| 47 |
-
"num_bits": 8,
|
| 48 |
-
"observer": "minmax",
|
| 49 |
-
"observer_kwargs": {},
|
| 50 |
-
"strategy": "channel",
|
| 51 |
-
"symmetric": true,
|
| 52 |
-
"type": "float"
|
| 53 |
-
}
|
| 54 |
-
}
|
| 55 |
-
},
|
| 56 |
-
"format": "float-quantized",
|
| 57 |
-
"global_compression_ratio": 1.5343121209820911,
|
| 58 |
-
"ignore": [
|
| 59 |
-
"lm_head"
|
| 60 |
-
],
|
| 61 |
-
"kv_cache_scheme": null,
|
| 62 |
-
"quant_method": "compressed-tensors",
|
| 63 |
-
"quantization_status": "compressed"
|
| 64 |
-
},
|
| 65 |
-
"rms_norm_eps": 1e-05,
|
| 66 |
-
"rope_scaling": {
|
| 67 |
-
"factor": 8.0,
|
| 68 |
-
"high_freq_factor": 4.0,
|
| 69 |
-
"low_freq_factor": 1.0,
|
| 70 |
-
"original_max_position_embeddings": 8192,
|
| 71 |
-
"rope_type": "llama3"
|
| 72 |
-
},
|
| 73 |
-
"rope_theta": 500000.0,
|
| 74 |
-
"tie_word_embeddings": false,
|
| 75 |
-
"torch_dtype": "bfloat16",
|
| 76 |
-
"transformers_version": "4.48.0",
|
| 77 |
-
"unsloth_fixed": true,
|
| 78 |
-
"use_cache": false,
|
| 79 |
-
"vocab_size": 128256
|
| 80 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|