Upload MistralForCausalLM

config.json

@@ -0,0 +1,26 @@
+{
+  "_name_or_path": "scb10x/typhoon-7b",
+  "architectures": [
+    "MistralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 10000.0,
+  "sliding_window": 4096,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.40.1",
+  "use_cache": true,
+  "vocab_size": 35219
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "transformers_version": "4.40.1"
+}

model-00001-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:472a2d1a0e29fb566b1e2c3812d9716b352469d73281eb2b7812c96728ba90d8
+size 979681984

model-00002-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f3fa12ddea7895a264bf7de7582dcca0ba30492b0aeeb3b72872629c9172e40
+size 872449040

model-00003-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6097ea18137876092a45f331059833de01d88f2c5128a0bafa5e1076fb93208
+size 872449040

model-00004-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d904d57236951cf8c0fe1717ac007911c419c913d7d28ae2f1edba8a0beb6071
+size 872449040

model-00005-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a3a837ff65538157b5a5fb5d3b3070b78df00614a313a3b7a8776ea52d1e3ec
+size 872449040

model-00006-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeef077af5cb25a235a674d4806fb33025206cb2f792872f00707383d543527a
+size 872449040

model-00007-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bf4298ecaa9d50de2eb777c3c390185de9ebfdb70bab22a03042fa2441bc733
+size 872449040

model-00008-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f429fad8171f320eea3a27c396100e89f76770810a65c87beb3a4379421fc3f1
+size 872449040

model-00009-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3dccb8d49b4998f26cae3f068e6304761bd23885ec698eebbc9f18923e75641
+size 872449040

model-00010-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8fcf01fdcafd0c9230d502bef66dad70e6c6657a6cd898c2560d322c753050e
+size 872449040

model-00011-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11f4a9f279ee8bf4bad2bdb23aa2cb56609abf7bd225506049b5662bca2fb660
+size 872449056

model-00012-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d5035d8a886850e07310b442b66b19b75f563fe1631db04a45d60bdb1833681
+size 872449048

model-00013-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc42e3cad208f6f0fa605d854267271fc320c197aded58fc36db4e887c45b75b
+size 872449048

model-00014-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db0db2674f9275b5ac316cba5d7e01d1c38b1f2ca2d6077e909f215213c517c1
+size 872449048

model-00015-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:496bba88731e6a4a127bd64d56c720034288510e754eb988aa833c058433df18
+size 872449048

model-00016-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3fcfe92e1621fda2285744a814838710f6f6a0fa2c136ca531a5ec352459577
+size 872449048

model-00017-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:649e3a4f49fb8d8be46a7a5b24391276f9ebfe08cc750f2bf474a56d46a72e92
+size 872449048

model-00018-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7da0c9fd376af6e5f6b57af1a2a8e226304a8b3f126f09b73ad3f886e0c3fda6
+size 872449048

model-00019-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:496fc2c5ca9468caf1375a1b33917ab61f5a87937fe0b86905b02d3373aa6a08
+size 872449048

model-00020-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a413d7571fd3f71390c8755aa437c638e752cccc82080f8718f927d6cdaaa330
+size 872449048

model-00021-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad5455648b3f20ed0e3a72d97ddc17379467f17bbe25482d2514d1aec697f94c
+size 872449048

model-00022-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:72e9c31a7f5c964d6dd81c84fa9ef598d4b57ace98358d8a0c51f1cba668487b
+size 872449048

model-00023-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2aab82e7b9f0ed52bf23737d239a07ef509efdc6db61dea4386faa1a1264505d
+size 872449048

model-00024-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfa66c70dd6b0e77c4e3747dc284062c58f52c4092fca7a3092c9a0b56226e9c
+size 872449048

model-00025-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef6d5be76bd2010bcf4ac1957d667ebf2eef778a34ea9ccfa003e0f16c44ff5c
+size 872449048

model-00026-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48d6d30f83d26fcf07b92ee431e7725a7f633766eb2166cef32569e9b6631673
+size 872449048

model-00027-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50ad08d0e3931cefefa6ea707c0c722b2bc040aa0939e01ef669319e7f919d3c
+size 872449048

model-00028-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:932069b4045a2cce5a17c9a2df2f86a52ef6afe44eded0f917ce6e949f341ebe
+size 872449048

model-00029-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc241a45a1cd2647c3cd17cad193498a8b5a83fcaef11c494f2e72e27e0b28e5
+size 872449048

model-00030-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:380077dbb0dc0b1f7f78036a5212221be2f1d544e87d5ee97439937270398e24
+size 872449048

model-00031-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fd516e8dca003a308f6979b85fe25da30b397a2e55f0899944ac25e9826906f
+size 872449048

model-00032-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae0e0b0574b897e747ec34705ca65180891cc49da9f6ebfbd70468bd4caaa8e0
+size 872449048

model-00033-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a52839dcd90f04d8d218c274a892ad0aebcefd5e3d14a465ba98a6a37bf313de
+size 469811768

model-00034-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58d018d8344e60954c8a7deb10f2eb7fd915c71925605662390080258b051d35
+size 577028224

model.safetensors.index.json

@@ -0,0 +1,298 @@
+{
+  "metadata": {
+    "total_size": 29072408576
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00034-of-00034.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00034.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00002-of-00034.safetensors",
+    "model.layers.0.mlp.down_proj.weight": "model-00002-of-00034.safetensors",
+    "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00034.safetensors",
+    "model.layers.0.mlp.up_proj.weight": "model-00002-of-00034.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00002-of-00034.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00034.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00034.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00034.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00034.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00003-of-00034.safetensors",
+    "model.layers.1.mlp.down_proj.weight": "model-00003-of-00034.safetensors",
+    "model.layers.1.mlp.gate_proj.weight": "model-00002-of-00034.safetensors",
+    "model.layers.1.mlp.up_proj.weight": "model-00003-of-00034.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00003-of-00034.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00002-of-00034.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00002-of-00034.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00002-of-00034.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00002-of-00034.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00012-of-00034.safetensors",
+    "model.layers.10.mlp.down_proj.weight": "model-00012-of-00034.safetensors",
+    "model.layers.10.mlp.gate_proj.weight": "model-00011-of-00034.safetensors",
+    "model.layers.10.mlp.up_proj.weight": "model-00012-of-00034.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00012-of-00034.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00011-of-00034.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00011-of-00034.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00011-of-00034.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00011-of-00034.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00013-of-00034.safetensors",
+    "model.layers.11.mlp.down_proj.weight": "model-00013-of-00034.safetensors",
+    "model.layers.11.mlp.gate_proj.weight": "model-00012-of-00034.safetensors",
+    "model.layers.11.mlp.up_proj.weight": "model-00013-of-00034.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00013-of-00034.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00012-of-00034.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00012-of-00034.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00012-of-00034.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00012-of-00034.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00014-of-00034.safetensors",
+    "model.layers.12.mlp.down_proj.weight": "model-00014-of-00034.safetensors",
+    "model.layers.12.mlp.gate_proj.weight": "model-00013-of-00034.safetensors",
+    "model.layers.12.mlp.up_proj.weight": "model-00014-of-00034.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00014-of-00034.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00013-of-00034.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00013-of-00034.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00013-of-00034.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00013-of-00034.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00015-of-00034.safetensors",
+    "model.layers.13.mlp.down_proj.weight": "model-00015-of-00034.safetensors",
+    "model.layers.13.mlp.gate_proj.weight": "model-00014-of-00034.safetensors",
+    "model.layers.13.mlp.up_proj.weight": "model-00015-of-00034.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00015-of-00034.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00014-of-00034.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00014-of-00034.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00014-of-00034.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00014-of-00034.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00016-of-00034.safetensors",
+    "model.layers.14.mlp.down_proj.weight": "model-00016-of-00034.safetensors",
+    "model.layers.14.mlp.gate_proj.weight": "model-00015-of-00034.safetensors",
+    "model.layers.14.mlp.up_proj.weight": "model-00016-of-00034.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00016-of-00034.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00015-of-00034.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00015-of-00034.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00015-of-00034.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00015-of-00034.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00017-of-00034.safetensors",
+    "model.layers.15.mlp.down_proj.weight": "model-00017-of-00034.safetensors",
+    "model.layers.15.mlp.gate_proj.weight": "model-00016-of-00034.safetensors",
+    "model.layers.15.mlp.up_proj.weight": "model-00017-of-00034.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00017-of-00034.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00016-of-00034.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00016-of-00034.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00016-of-00034.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00016-of-00034.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00018-of-00034.safetensors",
+    "model.layers.16.mlp.down_proj.weight": "model-00018-of-00034.safetensors",
+    "model.layers.16.mlp.gate_proj.weight": "model-00017-of-00034.safetensors",
+    "model.layers.16.mlp.up_proj.weight": "model-00018-of-00034.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00018-of-00034.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00017-of-00034.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00017-of-00034.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00017-of-00034.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00017-of-00034.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00019-of-00034.safetensors",
+    "model.layers.17.mlp.down_proj.weight": "model-00019-of-00034.safetensors",
+    "model.layers.17.mlp.gate_proj.weight": "model-00018-of-00034.safetensors",
+    "model.layers.17.mlp.up_proj.weight": "model-00019-of-00034.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00019-of-00034.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00018-of-00034.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00018-of-00034.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00018-of-00034.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00018-of-00034.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00020-of-00034.safetensors",
+    "model.layers.18.mlp.down_proj.weight": "model-00020-of-00034.safetensors",
+    "model.layers.18.mlp.gate_proj.weight": "model-00019-of-00034.safetensors",
+    "model.layers.18.mlp.up_proj.weight": "model-00020-of-00034.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00020-of-00034.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00019-of-00034.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00019-of-00034.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00019-of-00034.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00019-of-00034.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00021-of-00034.safetensors",
+    "model.layers.19.mlp.down_proj.weight": "model-00021-of-00034.safetensors",
+    "model.layers.19.mlp.gate_proj.weight": "model-00020-of-00034.safetensors",
+    "model.layers.19.mlp.up_proj.weight": "model-00021-of-00034.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00021-of-00034.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00020-of-00034.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00020-of-00034.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00020-of-00034.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00020-of-00034.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00004-of-00034.safetensors",
+    "model.layers.2.mlp.down_proj.weight": "model-00004-of-00034.safetensors",
+    "model.layers.2.mlp.gate_proj.weight": "model-00003-of-00034.safetensors",
+    "model.layers.2.mlp.up_proj.weight": "model-00004-of-00034.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00004-of-00034.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00003-of-00034.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00003-of-00034.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00003-of-00034.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00003-of-00034.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00022-of-00034.safetensors",
+    "model.layers.20.mlp.down_proj.weight": "model-00022-of-00034.safetensors",
+    "model.layers.20.mlp.gate_proj.weight": "model-00021-of-00034.safetensors",
+    "model.layers.20.mlp.up_proj.weight": "model-00022-of-00034.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00022-of-00034.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00021-of-00034.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00021-of-00034.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00021-of-00034.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00021-of-00034.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00023-of-00034.safetensors",
+    "model.layers.21.mlp.down_proj.weight": "model-00023-of-00034.safetensors",
+    "model.layers.21.mlp.gate_proj.weight": "model-00022-of-00034.safetensors",
+    "model.layers.21.mlp.up_proj.weight": "model-00023-of-00034.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00023-of-00034.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00022-of-00034.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00022-of-00034.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00022-of-00034.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00022-of-00034.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00024-of-00034.safetensors",
+    "model.layers.22.mlp.down_proj.weight": "model-00024-of-00034.safetensors",
+    "model.layers.22.mlp.gate_proj.weight": "model-00023-of-00034.safetensors",
+    "model.layers.22.mlp.up_proj.weight": "model-00024-of-00034.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00024-of-00034.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00023-of-00034.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00023-of-00034.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00023-of-00034.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00023-of-00034.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00025-of-00034.safetensors",
+    "model.layers.23.mlp.down_proj.weight": "model-00025-of-00034.safetensors",
+    "model.layers.23.mlp.gate_proj.weight": "model-00024-of-00034.safetensors",
+    "model.layers.23.mlp.up_proj.weight": "model-00025-of-00034.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00025-of-00034.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00024-of-00034.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00024-of-00034.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00024-of-00034.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00024-of-00034.safetensors",
+    "model.layers.24.input_layernorm.weight": "model-00026-of-00034.safetensors",
+    "model.layers.24.mlp.down_proj.weight": "model-00026-of-00034.safetensors",
+    "model.layers.24.mlp.gate_proj.weight": "model-00025-of-00034.safetensors",
+    "model.layers.24.mlp.up_proj.weight": "model-00026-of-00034.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "model-00026-of-00034.safetensors",
+    "model.layers.24.self_attn.k_proj.weight": "model-00025-of-00034.safetensors",
+    "model.layers.24.self_attn.o_proj.weight": "model-00025-of-00034.safetensors",
+    "model.layers.24.self_attn.q_proj.weight": "model-00025-of-00034.safetensors",
+    "model.layers.24.self_attn.v_proj.weight": "model-00025-of-00034.safetensors",
+    "model.layers.25.input_layernorm.weight": "model-00027-of-00034.safetensors",
+    "model.layers.25.mlp.down_proj.weight": "model-00027-of-00034.safetensors",
+    "model.layers.25.mlp.gate_proj.weight": "model-00026-of-00034.safetensors",
+    "model.layers.25.mlp.up_proj.weight": "model-00027-of-00034.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "model-00027-of-00034.safetensors",
+    "model.layers.25.self_attn.k_proj.weight": "model-00026-of-00034.safetensors",
+    "model.layers.25.self_attn.o_proj.weight": "model-00026-of-00034.safetensors",
+    "model.layers.25.self_attn.q_proj.weight": "model-00026-of-00034.safetensors",
+    "model.layers.25.self_attn.v_proj.weight": "model-00026-of-00034.safetensors",
+    "model.layers.26.input_layernorm.weight": "model-00028-of-00034.safetensors",
+    "model.layers.26.mlp.down_proj.weight": "model-00028-of-00034.safetensors",
+    "model.layers.26.mlp.gate_proj.weight": "model-00027-of-00034.safetensors",
+    "model.layers.26.mlp.up_proj.weight": "model-00028-of-00034.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "model-00028-of-00034.safetensors",
+    "model.layers.26.self_attn.k_proj.weight": "model-00027-of-00034.safetensors",
+    "model.layers.26.self_attn.o_proj.weight": "model-00027-of-00034.safetensors",
+    "model.layers.26.self_attn.q_proj.weight": "model-00027-of-00034.safetensors",
+    "model.layers.26.self_attn.v_proj.weight": "model-00027-of-00034.safetensors",
+    "model.layers.27.input_layernorm.weight": "model-00029-of-00034.safetensors",
+    "model.layers.27.mlp.down_proj.weight": "model-00029-of-00034.safetensors",
+    "model.layers.27.mlp.gate_proj.weight": "model-00028-of-00034.safetensors",
+    "model.layers.27.mlp.up_proj.weight": "model-00029-of-00034.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "model-00029-of-00034.safetensors",
+    "model.layers.27.self_attn.k_proj.weight": "model-00028-of-00034.safetensors",
+    "model.layers.27.self_attn.o_proj.weight": "model-00028-of-00034.safetensors",
+    "model.layers.27.self_attn.q_proj.weight": "model-00028-of-00034.safetensors",
+    "model.layers.27.self_attn.v_proj.weight": "model-00028-of-00034.safetensors",
+    "model.layers.28.input_layernorm.weight": "model-00030-of-00034.safetensors",
+    "model.layers.28.mlp.down_proj.weight": "model-00030-of-00034.safetensors",
+    "model.layers.28.mlp.gate_proj.weight": "model-00029-of-00034.safetensors",
+    "model.layers.28.mlp.up_proj.weight": "model-00030-of-00034.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "model-00030-of-00034.safetensors",
+    "model.layers.28.self_attn.k_proj.weight": "model-00029-of-00034.safetensors",
+    "model.layers.28.self_attn.o_proj.weight": "model-00029-of-00034.safetensors",
+    "model.layers.28.self_attn.q_proj.weight": "model-00029-of-00034.safetensors",
+    "model.layers.28.self_attn.v_proj.weight": "model-00029-of-00034.safetensors",
+    "model.layers.29.input_layernorm.weight": "model-00031-of-00034.safetensors",
+    "model.layers.29.mlp.down_proj.weight": "model-00031-of-00034.safetensors",
+    "model.layers.29.mlp.gate_proj.weight": "model-00030-of-00034.safetensors",
+    "model.layers.29.mlp.up_proj.weight": "model-00031-of-00034.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "model-00031-of-00034.safetensors",
+    "model.layers.29.self_attn.k_proj.weight": "model-00030-of-00034.safetensors",
+    "model.layers.29.self_attn.o_proj.weight": "model-00030-of-00034.safetensors",
+    "model.layers.29.self_attn.q_proj.weight": "model-00030-of-00034.safetensors",
+    "model.layers.29.self_attn.v_proj.weight": "model-00030-of-00034.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00005-of-00034.safetensors",
+    "model.layers.3.mlp.down_proj.weight": "model-00005-of-00034.safetensors",
+    "model.layers.3.mlp.gate_proj.weight": "model-00004-of-00034.safetensors",
+    "model.layers.3.mlp.up_proj.weight": "model-00005-of-00034.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00005-of-00034.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00004-of-00034.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00004-of-00034.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00004-of-00034.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00004-of-00034.safetensors",
+    "model.layers.30.input_layernorm.weight": "model-00032-of-00034.safetensors",
+    "model.layers.30.mlp.down_proj.weight": "model-00032-of-00034.safetensors",
+    "model.layers.30.mlp.gate_proj.weight": "model-00031-of-00034.safetensors",
+    "model.layers.30.mlp.up_proj.weight": "model-00032-of-00034.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "model-00032-of-00034.safetensors",
+    "model.layers.30.self_attn.k_proj.weight": "model-00031-of-00034.safetensors",
+    "model.layers.30.self_attn.o_proj.weight": "model-00031-of-00034.safetensors",
+    "model.layers.30.self_attn.q_proj.weight": "model-00031-of-00034.safetensors",
+    "model.layers.30.self_attn.v_proj.weight": "model-00031-of-00034.safetensors",
+    "model.layers.31.input_layernorm.weight": "model-00033-of-00034.safetensors",
+    "model.layers.31.mlp.down_proj.weight": "model-00033-of-00034.safetensors",
+    "model.layers.31.mlp.gate_proj.weight": "model-00032-of-00034.safetensors",
+    "model.layers.31.mlp.up_proj.weight": "model-00033-of-00034.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "model-00033-of-00034.safetensors",
+    "model.layers.31.self_attn.k_proj.weight": "model-00032-of-00034.safetensors",
+    "model.layers.31.self_attn.o_proj.weight": "model-00032-of-00034.safetensors",
+    "model.layers.31.self_attn.q_proj.weight": "model-00032-of-00034.safetensors",
+    "model.layers.31.self_attn.v_proj.weight": "model-00032-of-00034.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00006-of-00034.safetensors",
+    "model.layers.4.mlp.down_proj.weight": "model-00006-of-00034.safetensors",
+    "model.layers.4.mlp.gate_proj.weight": "model-00005-of-00034.safetensors",
+    "model.layers.4.mlp.up_proj.weight": "model-00006-of-00034.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00006-of-00034.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00005-of-00034.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00005-of-00034.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00005-of-00034.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00005-of-00034.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00007-of-00034.safetensors",
+    "model.layers.5.mlp.down_proj.weight": "model-00007-of-00034.safetensors",
+    "model.layers.5.mlp.gate_proj.weight": "model-00006-of-00034.safetensors",
+    "model.layers.5.mlp.up_proj.weight": "model-00007-of-00034.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00007-of-00034.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00006-of-00034.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00006-of-00034.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00006-of-00034.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00006-of-00034.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00008-of-00034.safetensors",
+    "model.layers.6.mlp.down_proj.weight": "model-00008-of-00034.safetensors",
+    "model.layers.6.mlp.gate_proj.weight": "model-00007-of-00034.safetensors",
+    "model.layers.6.mlp.up_proj.weight": "model-00008-of-00034.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00008-of-00034.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00007-of-00034.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00007-of-00034.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00007-of-00034.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00007-of-00034.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00009-of-00034.safetensors",
+    "model.layers.7.mlp.down_proj.weight": "model-00009-of-00034.safetensors",
+    "model.layers.7.mlp.gate_proj.weight": "model-00008-of-00034.safetensors",
+    "model.layers.7.mlp.up_proj.weight": "model-00009-of-00034.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00009-of-00034.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00008-of-00034.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00008-of-00034.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00008-of-00034.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00008-of-00034.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00010-of-00034.safetensors",
+    "model.layers.8.mlp.down_proj.weight": "model-00010-of-00034.safetensors",
+    "model.layers.8.mlp.gate_proj.weight": "model-00009-of-00034.safetensors",
+    "model.layers.8.mlp.up_proj.weight": "model-00010-of-00034.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00010-of-00034.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00009-of-00034.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00009-of-00034.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00009-of-00034.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00009-of-00034.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00011-of-00034.safetensors",
+    "model.layers.9.mlp.down_proj.weight": "model-00011-of-00034.safetensors",
+    "model.layers.9.mlp.gate_proj.weight": "model-00010-of-00034.safetensors",
+    "model.layers.9.mlp.up_proj.weight": "model-00011-of-00034.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00011-of-00034.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00010-of-00034.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00010-of-00034.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00010-of-00034.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00010-of-00034.safetensors",
+    "model.norm.weight": "model-00033-of-00034.safetensors"
+  }
+}