flexqwen-assin2 / config.json
mmcarpi's picture
Training in progress, epoch 1
9e30f1a verified
{
"architectures": [
"FlexQwenForSequenceClassification"
],
"attention_bias": false,
"attention_dropout": 0.0,
"auto_map": {
"AutoModel": "mmcarpi/flexqwen--qwen.FlexQwen",
"AutoModelForCausalLM": "mmcarpi/flexqwen--qwen.FlexQwenForCausalLM",
"AutoModelForSequenceClassification": "mmcarpi/flexqwen--qwen.FlexQwenForSequenceClassification"
},
"cls_token_id": 1,
"context_length": 4096,
"embedding_dim": 1024,
"head_dim": 128,
"hidden_act": "silu",
"hidden_dim": 2048,
"hidden_size": 4096,
"id2label": {
"0": "NONE",
"1": "ENTAILMENT"
},
"initializer_range": 0.02,
"intermediate_size": 22016,
"label2id": {
"ENTAILMENT": 1,
"NONE": 0
},
"max_position_embeddings": 32768,
"max_window_layers": 28,
"model_type": "qwen3",
"moe_hidden_dim": 512,
"moe_num_experts": 0,
"moe_num_experts_per_token": -1,
"num_attention_heads": 8,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"num_kv_groups": 8,
"pad_token_id": 3,
"qk_norm": true,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 10000,
"sliding_window": 4096,
"tie_word_embeddings": false,
"torch_dtype": "float32",
"transformers_version": "4.51.3",
"use_cache": true,
"use_sliding_window": false,
"vocab_size": 64000
}