Publish the export rather than the training checkpoint

config.json

@@ -0,0 +1,16 @@
+{
+  "bos_token_id": 2,
+  "dropout_prob": 0.1,
+  "eos_token_id": 3,
+  "hidden_size": 384,
+  "intermediate_size": 1152,
+  "label_smoothing": 0.05,
+  "layer_norm_eps": 1e-06,
+  "max_position_embeddings": 512,
+  "num_attention_heads": 6,
+  "num_decoder_layers": 6,
+  "num_encoder_layers": 6,
+  "pad_token_id": 0,
+  "unk_token_id": 1,
+  "vocab_size": 128
+}

export.stats.json

@@ -0,0 +1,31 @@
+{
+  "source": "artifacts/checkpoints/Juba-27M/juba_final.pt",
+  "source_contents": [
+    "config",
+    "model_state_dict",
+    "val_acc"
+  ],
+  "dropped_state": [],
+  "dropped_tied": [
+    "lm_head.weight"
+  ],
+  "tensors": 145,
+  "parameters": 26901888,
+  "files": [
+    {
+      "name": "README.md",
+      "bytes": 10616,
+      "sha256": "17a73ad594732e0a33e6890acef29a8ceb251fa8698b38ed7eef70201af36654"
+    },
+    {
+      "name": "config.json",
+      "bytes": 352,
+      "sha256": "4b94ed28a330e30a4591eaf547ad68e151b096f366963bbfe8427d5030c672cf"
+    },
+    {
+      "name": "model.safetensors",
+      "bytes": 107622888,
+      "sha256": "d6ea5fd03a6fa54a61a4f8a7956d7433dbb80d24bb90e9f59c4a472360b66389"
+    }
+  ]
+}

juba_final.pt → model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8af8abc2d805e499f89556b84056a132f41a47e393f476e015f81736a9debd60
-size 107666274
+oid sha256:d6ea5fd03a6fa54a61a4f8a7956d7433dbb80d24bb90e9f59c4a472360b66389
+size 107622888