internlm3-8b-instruct

internlm/internlm3-8b-instruct/bert4torch_config.json

@@ -0,0 +1,33 @@
+{
+    "model": "llama",
+    "template": "apply_chat_template",
+    "bias": false,
+    "bos_token_id": 1,
+    "eos_token_id": 2,
+    "pad_token_id": 2,
+    "hidden_act": "silu",
+    "hidden_size": 4096,
+    "initializer_range": 0.02,
+    "intermediate_size": 10240,
+    "max_position_embeddings": 32768,
+    "num_attention_heads": 32,
+    "num_hidden_layers": 48,
+    "num_key_value_heads": 2,
+    "layer_norm_eps": 1e-05,
+    "rope_scaling": {
+      "type": "dynamic",
+      "factor": 6.0
+    },
+    "rope_theta": 50000000,
+    "tie_word_embeddings": false,
+    "torch_dtype": "bfloat16",
+    "vocab_size": 128512,
+    "segment_vocab_size": 0,
+    "skip_init": true,
+    "rope_rank": "updown",
+    "generation_config": {
+      "tokenizer_config":  {"skip_special_tokens": true, "add_special_tokens": false}, 
+      "eos_token_id": [2, 128131], 
+      "max_length": 32768
+    }
+  }