Adding `safetensors` variant of this model

model-00001-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b377bc86145a3e1a002b9d5b9015f0a993a12ede5d19ec877fdd456ddf042ec1
+size 197448232

model-00002-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdd682f3a9bbecf62d11a8041ab4c9d3c74cb4a15bf28c3b992b385bf3a7a707
+size 205810396

model-00003-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:351f64db124cfaf8c441c274e5693d1d2b8851726ea211795d661921e5c9234d
+size 107100908

model-00004-of-00004.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2750d01044b74bf2416122b1801b467d8de320de53cc1535e0d357062692fc7d
+size 154389632

model.safetensors.index.json

@@ -0,0 +1,180 @@
+{
+    "metadata": {
+        "total_size": 653721648.0
+    },
+    "weight_map": {
+        "lm_head.weight": "model-00004-of-00004.safetensors",
+        "transformer.h.0.attn.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.0.attn.c_attn.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.0.attn.c_attn.weight": "model-00001-of-00004.safetensors",
+        "transformer.h.0.attn.c_proj.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.0.attn.c_proj.weight": "model-00001-of-00004.safetensors",
+        "transformer.h.0.attn.masked_bias": "model-00001-of-00004.safetensors",
+        "transformer.h.0.ln_1.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.0.ln_1.weight": "model-00001-of-00004.safetensors",
+        "transformer.h.0.ln_2.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.0.ln_2.weight": "model-00001-of-00004.safetensors",
+        "transformer.h.0.mlp.c_fc.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.0.mlp.c_fc.weight": "model-00001-of-00004.safetensors",
+        "transformer.h.0.mlp.c_proj.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.0.mlp.c_proj.weight": "model-00001-of-00004.safetensors",
+        "transformer.h.1.attn.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.1.attn.c_attn.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.1.attn.c_attn.weight": "model-00001-of-00004.safetensors",
+        "transformer.h.1.attn.c_proj.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.1.attn.c_proj.weight": "model-00001-of-00004.safetensors",
+        "transformer.h.1.attn.masked_bias": "model-00001-of-00004.safetensors",
+        "transformer.h.1.ln_1.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.1.ln_1.weight": "model-00001-of-00004.safetensors",
+        "transformer.h.1.ln_2.bias": "model-00001-of-00004.safetensors",
+        "transformer.h.1.ln_2.weight": "model-00001-of-00004.safetensors",
+        "transformer.h.1.mlp.c_fc.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.1.mlp.c_fc.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.1.mlp.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.1.mlp.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.10.attn.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.10.attn.c_attn.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.10.attn.c_attn.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.10.attn.c_proj.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.10.attn.c_proj.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.10.attn.masked_bias": "model-00003-of-00004.safetensors",
+        "transformer.h.10.ln_1.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.10.ln_1.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.10.ln_2.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.10.ln_2.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.10.mlp.c_fc.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.10.mlp.c_fc.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.10.mlp.c_proj.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.10.mlp.c_proj.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.11.attn.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.11.attn.c_attn.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.11.attn.c_attn.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.11.attn.c_proj.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.11.attn.c_proj.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.11.attn.masked_bias": "model-00003-of-00004.safetensors",
+        "transformer.h.11.ln_1.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.11.ln_1.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.11.ln_2.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.11.ln_2.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.11.mlp.c_fc.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.11.mlp.c_fc.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.11.mlp.c_proj.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.11.mlp.c_proj.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.2.attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.2.attn.c_attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.2.attn.c_attn.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.2.attn.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.2.attn.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.2.attn.masked_bias": "model-00002-of-00004.safetensors",
+        "transformer.h.2.ln_1.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.2.ln_1.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.2.ln_2.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.2.ln_2.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.2.mlp.c_fc.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.2.mlp.c_fc.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.2.mlp.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.2.mlp.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.3.attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.3.attn.c_attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.3.attn.c_attn.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.3.attn.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.3.attn.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.3.attn.masked_bias": "model-00002-of-00004.safetensors",
+        "transformer.h.3.ln_1.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.3.ln_1.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.3.ln_2.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.3.ln_2.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.3.mlp.c_fc.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.3.mlp.c_fc.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.3.mlp.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.3.mlp.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.4.attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.4.attn.c_attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.4.attn.c_attn.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.4.attn.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.4.attn.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.4.attn.masked_bias": "model-00002-of-00004.safetensors",
+        "transformer.h.4.ln_1.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.4.ln_1.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.4.ln_2.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.4.ln_2.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.4.mlp.c_fc.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.4.mlp.c_fc.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.4.mlp.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.4.mlp.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.5.attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.5.attn.c_attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.5.attn.c_attn.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.5.attn.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.5.attn.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.5.attn.masked_bias": "model-00002-of-00004.safetensors",
+        "transformer.h.5.ln_1.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.5.ln_1.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.5.ln_2.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.5.ln_2.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.5.mlp.c_fc.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.5.mlp.c_fc.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.5.mlp.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.5.mlp.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.6.attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.6.attn.c_attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.6.attn.c_attn.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.6.attn.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.6.attn.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.6.attn.masked_bias": "model-00002-of-00004.safetensors",
+        "transformer.h.6.ln_1.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.6.ln_1.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.6.ln_2.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.6.ln_2.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.6.mlp.c_fc.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.6.mlp.c_fc.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.6.mlp.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.6.mlp.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.7.attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.7.attn.c_attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.7.attn.c_attn.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.7.attn.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.7.attn.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.7.attn.masked_bias": "model-00002-of-00004.safetensors",
+        "transformer.h.7.ln_1.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.7.ln_1.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.7.ln_2.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.7.ln_2.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.7.mlp.c_fc.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.7.mlp.c_fc.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.7.mlp.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.7.mlp.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.8.attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.8.attn.c_attn.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.8.attn.c_attn.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.8.attn.c_proj.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.8.attn.c_proj.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.8.attn.masked_bias": "model-00002-of-00004.safetensors",
+        "transformer.h.8.ln_1.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.8.ln_1.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.8.ln_2.bias": "model-00002-of-00004.safetensors",
+        "transformer.h.8.ln_2.weight": "model-00002-of-00004.safetensors",
+        "transformer.h.8.mlp.c_fc.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.8.mlp.c_fc.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.8.mlp.c_proj.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.8.mlp.c_proj.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.9.attn.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.9.attn.c_attn.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.9.attn.c_attn.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.9.attn.c_proj.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.9.attn.c_proj.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.9.attn.masked_bias": "model-00003-of-00004.safetensors",
+        "transformer.h.9.ln_1.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.9.ln_1.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.9.ln_2.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.9.ln_2.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.9.mlp.c_fc.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.9.mlp.c_fc.weight": "model-00003-of-00004.safetensors",
+        "transformer.h.9.mlp.c_proj.bias": "model-00003-of-00004.safetensors",
+        "transformer.h.9.mlp.c_proj.weight": "model-00003-of-00004.safetensors",
+        "transformer.ln_f.bias": "model-00003-of-00004.safetensors",
+        "transformer.ln_f.weight": "model-00003-of-00004.safetensors",
+        "transformer.wpe.weight": "model-00001-of-00004.safetensors",
+        "transformer.wte.weight": "model-00001-of-00004.safetensors"
+    }
+}