Upload folder using huggingface_hub

debug.log

@@ -0,0 +1,3 @@
+name shakespeare_64x4 | device cuda | compile True | data_dir data/shakespeare | should_randomize True | log_interval 10 | eval_interval 250 | eval_steps 100 | batch_size 128 | gradient_accumulation_steps 1 | learning_rate 0.001 | warmup_steps 1000 | max_steps 10000 | decay_lr True | min_lr 1e-05 | weight_decay 0.1 | grad_clip 1.0 | gpt_config {'name': 'ascii_64x4', 'device': device(type='cuda'), 'compile': True, 'block_size': 128, 'vocab_size': 128, 'n_layer': 4, 'n_head': 4, 'n_embd': 64}
+name shakespeare_64x4 | device cuda | compile True | data_dir data/shakespeare | should_randomize True | log_interval 10 | eval_interval 250 | eval_steps 100 | batch_size 128 | gradient_accumulation_steps 1 | learning_rate 0.001 | warmup_steps 1000 | max_steps 10000 | decay_lr True | min_lr 1e-05 | weight_decay 0.1 | grad_clip 1.0 | gpt_config {'name': 'ascii_64x4', 'device': device(type='cuda'), 'compile': True, 'block_size': 128, 'vocab_size': 128, 'n_layer': 4, 'n_head': 4, 'n_embd': 64}
+name shakespeare_64x4 | device cuda | compile True | data_dir data/shakespeare | should_randomize True | log_interval 10 | eval_interval 250 | eval_steps 100 | batch_size 128 | gradient_accumulation_steps 1 | learning_rate 0.001 | warmup_steps 1000 | max_steps 10000 | decay_lr True | min_lr 1e-05 | weight_decay 0.1 | grad_clip 1.0 | gpt_config {'name': 'ascii_64x4', 'device': device(type='cuda'), 'compile': True, 'block_size': 128, 'vocab_size': 128, 'n_layer': 4, 'n_head': 4, 'n_embd': 64}

eval.log

@@ -0,0 +1,83 @@
+type eval | step 0 | loss 4.8501 | checkpoint False
+type eval | step 0 | loss 4.8489 | checkpoint False
+type eval | step 250 | loss 3.1139 | checkpoint True
+type eval | step 500 | loss 2.5083 | checkpoint True
+type eval | step 750 | loss 2.3826 | checkpoint True
+type eval | step 1000 | loss 2.2400 | checkpoint True
+type eval | step 1250 | loss 2.0680 | checkpoint True
+type eval | step 1500 | loss 1.9792 | checkpoint True
+type eval | step 1750 | loss 1.9354 | checkpoint True
+type eval | step 2000 | loss 1.8793 | checkpoint True
+type eval | step 2250 | loss 1.8414 | checkpoint True
+type eval | step 2500 | loss 1.8133 | checkpoint True
+type eval | step 2750 | loss 1.7787 | checkpoint True
+type eval | step 3000 | loss 1.7515 | checkpoint True
+type eval | step 3250 | loss 1.7513 | checkpoint True
+type eval | step 3500 | loss 1.7423 | checkpoint True
+type eval | step 3750 | loss 1.7245 | checkpoint True
+type eval | step 4000 | loss 1.7110 | checkpoint True
+type eval | step 4250 | loss 1.7084 | checkpoint True
+type eval | step 4500 | loss 1.6869 | checkpoint True
+type eval | step 4750 | loss 1.6772 | checkpoint True
+type eval | step 5000 | loss 1.6680 | checkpoint True
+type eval | step 5250 | loss 1.6416 | checkpoint True
+type eval | step 5500 | loss 1.6447 | checkpoint False
+type eval | step 5750 | loss 1.6367 | checkpoint True
+type eval | step 6000 | loss 1.6424 | checkpoint False
+type eval | step 6250 | loss 1.6370 | checkpoint False
+type eval | step 6500 | loss 1.6343 | checkpoint True
+type eval | step 6750 | loss 1.6303 | checkpoint True
+type eval | step 7000 | loss 1.6229 | checkpoint True
+type eval | step 7250 | loss 1.6121 | checkpoint True
+type eval | step 7500 | loss 1.6108 | checkpoint True
+type eval | step 7750 | loss 1.6004 | checkpoint True
+type eval | step 8000 | loss 1.5971 | checkpoint True
+type eval | step 8250 | loss 1.5931 | checkpoint True
+type eval | step 8500 | loss 1.5917 | checkpoint True
+type eval | step 8750 | loss 1.5924 | checkpoint False
+type eval | step 9000 | loss 1.5881 | checkpoint True
+type eval | step 9250 | loss 1.5856 | checkpoint True
+type eval | step 9500 | loss 1.5845 | checkpoint True
+type eval | step 9750 | loss 1.5839 | checkpoint True
+type eval | step 10000 | loss 1.5833 | checkpoint True
+type eval | step 0 | loss 4.8544 | checkpoint False
+type eval | step 250 | loss 3.0956 | checkpoint True
+type eval | step 500 | loss 2.4943 | checkpoint True
+type eval | step 750 | loss 2.3488 | checkpoint True
+type eval | step 1000 | loss 2.1962 | checkpoint True
+type eval | step 1250 | loss 2.0515 | checkpoint True
+type eval | step 1500 | loss 1.9533 | checkpoint True
+type eval | step 1750 | loss 1.9070 | checkpoint True
+type eval | step 2000 | loss 1.8539 | checkpoint True
+type eval | step 2250 | loss 1.8142 | checkpoint True
+type eval | step 2500 | loss 1.7810 | checkpoint True
+type eval | step 2750 | loss 1.7395 | checkpoint True
+type eval | step 3000 | loss 1.7162 | checkpoint True
+type eval | step 3250 | loss 1.7088 | checkpoint True
+type eval | step 3500 | loss 1.6995 | checkpoint True
+type eval | step 3750 | loss 1.6880 | checkpoint True
+type eval | step 4000 | loss 1.6766 | checkpoint True
+type eval | step 4250 | loss 1.6753 | checkpoint True
+type eval | step 4500 | loss 1.6584 | checkpoint True
+type eval | step 4750 | loss 1.6419 | checkpoint True
+type eval | step 5000 | loss 1.6300 | checkpoint True
+type eval | step 5250 | loss 1.6117 | checkpoint True
+type eval | step 5500 | loss 1.6086 | checkpoint True
+type eval | step 5750 | loss 1.6152 | checkpoint False
+type eval | step 6000 | loss 1.6096 | checkpoint False
+type eval | step 6250 | loss 1.6190 | checkpoint False
+type eval | step 6500 | loss 1.6118 | checkpoint False
+type eval | step 6750 | loss 1.6055 | checkpoint True
+type eval | step 7000 | loss 1.6066 | checkpoint False
+type eval | step 7250 | loss 1.6080 | checkpoint False
+type eval | step 7500 | loss 1.6004 | checkpoint True
+type eval | step 7750 | loss 1.5821 | checkpoint True
+type eval | step 8000 | loss 1.5792 | checkpoint True
+type eval | step 8250 | loss 1.5776 | checkpoint True
+type eval | step 8500 | loss 1.5804 | checkpoint False
+type eval | step 8750 | loss 1.5805 | checkpoint False
+type eval | step 9000 | loss 1.5763 | checkpoint True
+type eval | step 9250 | loss 1.5718 | checkpoint True
+type eval | step 9500 | loss 1.5695 | checkpoint True
+type eval | step 9750 | loss 1.5686 | checkpoint True
+type eval | step 10000 | loss 1.5683 | checkpoint True

model.json

@@ -0,0 +1 @@
+{"block_size": 128, "vocab_size": 128, "n_layer": 4, "n_head": 4, "n_embd": 64}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b7a990722c440fdbd1e79a78b08a6f65f74fe96acfa49e5ef4cd5c8936a3c85
+size 870720