Spaces:
Running
Running
Update results
Browse files- evaluate.py +3 -3
- results.json +63 -1
evaluate.py
CHANGED
|
@@ -9,7 +9,7 @@ import json
|
|
| 9 |
device = "cuda"
|
| 10 |
test_indices_length = 10
|
| 11 |
|
| 12 |
-
models = ["distributed/optimized-gpt2-250m", "distributed/gpt2-94m"]
|
| 13 |
|
| 14 |
with open('results.json', 'r') as file:
|
| 15 |
results = json.load(file)
|
|
@@ -79,5 +79,5 @@ for model_name in models:
|
|
| 79 |
results[model_name][str(epoch)] = [average_loss]
|
| 80 |
print(f"Epoch: {epoch} Average Loss: {average_loss:.2f}")
|
| 81 |
|
| 82 |
-
with open("results.json", "w") as outfile:
|
| 83 |
-
|
|
|
|
| 9 |
device = "cuda"
|
| 10 |
test_indices_length = 10
|
| 11 |
|
| 12 |
+
models = ["distributed/optimized-gpt2-250m", "distributed/optimized-gpt2-250m-v0.1.1", "distributed/gpt2-94m"]
|
| 13 |
|
| 14 |
with open('results.json', 'r') as file:
|
| 15 |
results = json.load(file)
|
|
|
|
| 79 |
results[model_name][str(epoch)] = [average_loss]
|
| 80 |
print(f"Epoch: {epoch} Average Loss: {average_loss:.2f}")
|
| 81 |
|
| 82 |
+
with open("results.json", "w") as outfile:
|
| 83 |
+
json.dump(results, outfile, indent = 4)
|
results.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"distributed/optimized-gpt2-250m": {
|
| 3 |
"0": [
|
| 4 |
11.042416954040528
|
| 5 |
],
|
|
@@ -1703,5 +1703,67 @@
|
|
| 1703 |
"256": [
|
| 1704 |
7.2350592613220215
|
| 1705 |
]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1706 |
}
|
| 1707 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"distributed/optimized-gpt2-250m-v0.1.1": {
|
| 3 |
"0": [
|
| 4 |
11.042416954040528
|
| 5 |
],
|
|
|
|
| 1703 |
"256": [
|
| 1704 |
7.2350592613220215
|
| 1705 |
]
|
| 1706 |
+
},
|
| 1707 |
+
"distributed/optimized-gpt2-250m": {
|
| 1708 |
+
"0": [
|
| 1709 |
+
10.9341459274292
|
| 1710 |
+
],
|
| 1711 |
+
"1": [
|
| 1712 |
+
10.934042739868165
|
| 1713 |
+
],
|
| 1714 |
+
"2": [
|
| 1715 |
+
10.944120025634765
|
| 1716 |
+
],
|
| 1717 |
+
"3": [
|
| 1718 |
+
10.955198446909586
|
| 1719 |
+
],
|
| 1720 |
+
"4": [
|
| 1721 |
+
10.947817611694337
|
| 1722 |
+
],
|
| 1723 |
+
"5": [
|
| 1724 |
+
10.95099925994873
|
| 1725 |
+
],
|
| 1726 |
+
"6": [
|
| 1727 |
+
10.918320020039877
|
| 1728 |
+
],
|
| 1729 |
+
"7": [
|
| 1730 |
+
10.917640924453735
|
| 1731 |
+
],
|
| 1732 |
+
"8": [
|
| 1733 |
+
10.934884071350098
|
| 1734 |
+
],
|
| 1735 |
+
"9": [
|
| 1736 |
+
10.920173645019531
|
| 1737 |
+
],
|
| 1738 |
+
"10": [
|
| 1739 |
+
10.913983345031738
|
| 1740 |
+
],
|
| 1741 |
+
"11": [
|
| 1742 |
+
10.854215621948242
|
| 1743 |
+
],
|
| 1744 |
+
"12": [
|
| 1745 |
+
10.904011011123657
|
| 1746 |
+
],
|
| 1747 |
+
"13": [
|
| 1748 |
+
10.89674186706543
|
| 1749 |
+
],
|
| 1750 |
+
"14": [
|
| 1751 |
+
10.912863731384277
|
| 1752 |
+
],
|
| 1753 |
+
"15": [
|
| 1754 |
+
10.8756103515625
|
| 1755 |
+
],
|
| 1756 |
+
"16": [
|
| 1757 |
+
10.82695255279541
|
| 1758 |
+
],
|
| 1759 |
+
"17": [
|
| 1760 |
+
10.846389055252075
|
| 1761 |
+
],
|
| 1762 |
+
"18": [
|
| 1763 |
+
10.82868251800537
|
| 1764 |
+
],
|
| 1765 |
+
"19": [
|
| 1766 |
+
10.803816986083984
|
| 1767 |
+
]
|
| 1768 |
}
|
| 1769 |
}
|