Upload 2 files

config.json

@@ -3,55 +3,49 @@
   "model_type": "midi_model",
   "n_embd": 1024,
   "net_config": {
-    "attention_bias": false,
     "attention_dropout": 0.0,
-    "attn_logit_softcapping": null,
-    "cache_implementation": "hybrid",
-    "final_logit_softcapping": null,
-    "head_dim": 256,
-    "hidden_activation": "gelu_pytorch_tanh",
+    "embd_pdrop": 0.0,
+    "hidden_act": "silu",
     "hidden_size": 1024,
     "initializer_range": 0.02,
     "intermediate_size": 4096,
-    "max_position_embeddings": 4096,
-    "model_type": "gemma3_text",
+    "max_position_embeddings": 2048,
+    "model_type": "phi3",
     "num_attention_heads": 16,
     "num_hidden_layers": 12,
-    "num_key_value_heads": 8,
-    "query_pre_attn_scalar": 256,
-    "rms_norm_eps": 1e-06,
-    "rope_local_base_freq": 10000.0,
+    "num_key_value_heads": 16,
+    "original_max_position_embeddings": 4096,
+    "pad_token_id": 0,
+    "partial_rotary_factor": 1.0,
+    "resid_pdrop": 0.0,
+    "rms_norm_eps": 1e-05,
     "rope_scaling": null,
-    "rope_theta": 1000000.0,
-    "sliding_window": 512,
-    "sliding_window_pattern": 6,
-    "use_cache": true,
+    "rope_theta": 10000.0,
+    "sliding_window": null,
+    "use_cache": false,
     "vocab_size": 3406
   },
   "net_token_config": {
-    "attention_bias": false,
     "attention_dropout": 0.0,
-    "attn_logit_softcapping": null,
-    "cache_implementation": "hybrid",
-    "final_logit_softcapping": null,
-    "head_dim": 256,
-    "hidden_activation": "gelu_pytorch_tanh",
+    "embd_pdrop": 0.0,
+    "hidden_act": "silu",
     "hidden_size": 1024,
     "initializer_range": 0.02,
     "intermediate_size": 1024,
-    "max_position_embeddings": 4096,
-    "model_type": "gemma3_text",
+    "max_position_embeddings": 2048,
+    "model_type": "phi3",
     "num_attention_heads": 4,
     "num_hidden_layers": 3,
-    "num_key_value_heads": 1,
-    "query_pre_attn_scalar": 256,
-    "rms_norm_eps": 1e-06,
-    "rope_local_base_freq": 10000.0,
+    "num_key_value_heads": 4,
+    "original_max_position_embeddings": 4096,
+    "pad_token_id": 0,
+    "partial_rotary_factor": 1.0,
+    "resid_pdrop": 0.0,
+    "rms_norm_eps": 1e-05,
     "rope_scaling": null,
-    "rope_theta": 1000000.0,
-    "sliding_window": 512,
-    "sliding_window_pattern": 6,
-    "use_cache": true,
+    "rope_theta": 10000.0,
+    "sliding_window": null,
+    "use_cache": false,
     "vocab_size": 3406
   },
   "tokenizer": {

model.ckpt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d54e36c5a370b15f2cd131d6b56613f1ab9fc1ff2ab4695d4e82b5156c11f362
-size 1979207421
+oid sha256:eaca584a7e2d7beab42e783dcf4d9f845957465484f71f4ce584cfb017710ae1
+size 1403181381