| { | |
| "architectures": [ | |
| "IQuestCoderForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "auto_map": { | |
| "AutoConfig": "configuration_iquestcoder.IQuestCoderConfig", | |
| "AutoModel": "modeling_iquestcoder.IQuestCoderModel", | |
| "AutoModelForCausalLM": "modeling_iquestcoder.IQuestCoderForCausalLM", | |
| "AutoModelForQuestionAnswering": "modeling_iquestcoder.IQuestCoderForQuestionAnswering", | |
| "AutoModelForSequenceClassification": "modeling_iquestcoder.IQuestCoderForSequenceClassification", | |
| "AutoModelForTokenClassification": "modeling_iquestcoder.IQuestCoderForTokenClassification" | |
| }, | |
| "bos_token_id": 1, | |
| "clip_qkv": null, | |
| "eos_token_id": [ | |
| 2, | |
| 75864, | |
| 75869 | |
| ], | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 5120, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 27648, | |
| "max_position_embeddings": 131072, | |
| "max_window_layers": 0, | |
| "mlp_bias": false, | |
| "model_type": "iquestcoder", | |
| "num_attention_heads": 40, | |
| "num_hidden_layers": 64, | |
| "num_key_value_heads": 8, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": null, | |
| "rope_theta": 500000.0, | |
| "sliding_window": null, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.52.0", | |
| "use_cache": false, | |
| "use_sliding_window": false, | |
| "vocab_size": 76800, | |
| "compression_config": { | |
| "config_groups": { | |
| "group_0": { | |
| "targets": [ | |
| "Linear" | |
| ], | |
| "input_activations": { | |
| "dynamic": true, | |
| "group_size": null, | |
| "num_bits": 8, | |
| "observer": "minmax", | |
| "observer_kwargs": {}, | |
| "strategy": "token", | |
| "symmetric": true, | |
| "type": "float" | |
| }, | |
| "weights": { | |
| "dynamic": false, | |
| "group_size": null, | |
| "num_bits": 8, | |
| "observer": "minmax", | |
| "observer_kwargs": {}, | |
| "strategy": "channel", | |
| "symmetric": true, | |
| "type": "float" | |
| } | |
| } | |
| }, | |
| "format": "float-quantized", | |
| "ignore": [ | |
| "lm_head" | |
| ], | |
| "quant_method": "compressed-tensors" | |
| } | |
| } |