Upload LlamaForCausalLM

config.json

@@ -0,0 +1,28 @@
+{
+  "_name_or_path": "meta-llama/Meta-Llama-3-8B",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 128000,
+  "eos_token_id": 128001,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 8192,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 500000.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.40.1",
+  "use_cache": true,
+  "vocab_size": 128256
+}

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "bos_token_id": 128000,
+  "do_sample": true,
+  "eos_token_id": 128001,
+  "max_length": 4096,
+  "temperature": 0.6,
+  "top_p": 0.9,
+  "transformers_version": "4.40.1"
+}

model-00001-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a0b52a19ff14c694275e8a0b6c78037cce9e9193069c1d3d5ec22498389ea0c
+size 2101346440

model-00002-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8910c492c86c978cee67dc9a66b9a0a7b90aa5ed164326dca639626ba1f6556a
+size 973112680

model-00003-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3fc82c2ee6a6d32971f7d934586b621403bbc98ce8906e4c8cdc1e351efd7f2
+size 939558024

model-00004-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d16f2a57343b79d6c509c655ce035bdaf79f4610174e67abe273f65089cd0b65
+size 872449040

model-00005-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e98543d891bd9bc0481b03a54800598ce0550420a3c2d02f043217f3cd6dea31
+size 872449040

model-00006-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f96b1900767d80abaf8dfc59dd0d1e78db1aea201bfebea77a4c45575244270
+size 872449040

model-00007-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:783d2a002ba7f710ed3971b9c3198af6fc254167b3bd560c759e080aaad8af35
+size 872449040

model-00008-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48b0793923c12fa0b1e0077f653f4893b2e51b633dedb534c5ccd020fdeb4c42
+size 872449040

model-00009-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2bcb4ec46795673046cb58e551685c3f60dadebdfd9dcf3a9ceda01178020c12
+size 872449040

model-00010-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fe2a65732c0d22f6bf4fde198a92c83c0789c5874145b2699c68bac3c3fcc99
+size 872449040

model-00011-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cecc3afff6672c729e7673d5935b63f6eead9ac8965388df4e5865755bb25d7
+size 872449048

model-00012-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08e880b4b60c84297b0153876226813085b184aeca06317e3907d31d4d8efe9f
+size 872449048

model-00013-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3bef7ac4c0cf377efe9a71758b74ca681bf69fc8e6ee60b10c162a1d001f3c5
+size 872449048

model-00014-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9238ff577b606c4cf8d42caeb9684a84fcd0f268531b127a2f2cfb6cf6861efb
+size 872449048

model-00015-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6242f328c5268d7ceb90331e417d9ea01df9d17e1a171a3086c84f9ded8112c
+size 872449048

model-00016-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b78914969bed794d4a8dd0b85d2e293244f50c67af7addaebc6c313e4ac67fe7
+size 872449048

model-00017-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82ef5fe34aafaf9bf8431ad030da02236ab6a4eda143e2be93c80219d4c6b706
+size 872449048

model-00018-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9c5e43294f111aacbe05f28dd7c9731d82e83f25758cb27a3179b6a1edc079e
+size 872449048

model-00019-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40d3faa2834f4de0c2c7d4f49445272baf0c5e5aa16261f4678354a9f425aa2b
+size 872449048

model-00020-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:648a4565f79652f5dae1bc6571fbfa1c6d45753755c3b0a9ba463e397398894c
+size 872449048

model-00021-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0412f8937ec51eff27328ab749d8ebb8f6f76b2d4f6a9f4c6bb78279afc5faf7
+size 872449048

model-00022-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7c157ad4bde27215394a7833603d365aef7081e962c4de50e149669623af85b
+size 872449048

model-00023-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:040e49c2770a81ae4007fda7c1a28310667619f36e60ce8fca817a76e39472e3
+size 872449048

model-00024-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bac54f4d8ece671a99488b3c33e32ebfef4b497a173906020b27f160fe9b6d58
+size 872449048

model-00025-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14a4d8c9b54cd09d808f41ebde2551cf44021874404f6088e743489a72e54df2
+size 872449048

model-00026-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61224f4c832e0c7b8242163fa92c79ad08d1f9373b44c402055f9075d2921cc4
+size 872449048

model-00027-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:deaed345aacb0688a0e3db06adc06d1951b7bb494a3096aee9b0a623f9189af3
+size 872449048

model-00028-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6b819e332df1dba5977dbb2f54aba908de74c7bbea81b441dfa2b4575fa23f4
+size 872449048

model-00029-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96027bb9e69ded8e4c950c9b00d17e1f8392fb1cb80a713fc91965fe17c60be0
+size 872449048

model-00030-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ab670c023e3a2fceec97e84624b1f194510da3ca81324228f6693e3bfe0db2e
+size 872449048

model-00031-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aade4fc60dffea5082b80c738890867e8040939650c1c16690eef2dfa7ebe978
+size 872449048

model-00032-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4547ef2febb0c1b8176ac691990b01cd1bcf934e73cbc91b9e8afef90c6016c3
+size 872449048

model-00033-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e3ba679a425795f03ebafc26e4c52ee66d4802e9fd03b3b223c75a33815f58d
+size 704692904

model-00034-of-00034.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5357ab75c2e5bf52265120e14cf0f545156b7a3ea99808ac135562b30578ecad
+size 2101346432

model.safetensors.index.json

@@ -0,0 +1,298 @@
+{
+  "metadata": {
+    "total_size": 32121044992
+  },
+  "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-00002-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-00002-of-00034.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00002-of-00034.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00002-of-00034.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00002-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-00003-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-00003-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-00012-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-00013-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-00014-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-00015-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-00016-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-00017-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-00018-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-00019-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-00020-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-00021-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-00004-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-00022-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-00023-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-00024-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-00025-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-00026-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-00027-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-00028-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-00029-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-00030-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-00031-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-00005-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-00032-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-00033-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-00006-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-00007-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-00008-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-00009-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-00010-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-00011-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"
+  }
+}