Geralt-Targaryen
commited on
Commit
•
8cb0727
1
Parent(s):
e5a577a
v5
Browse files- README.md +15 -9
- config.json +1 -1
- merges.txt +0 -0
- pytorch_model.bin +1 -1
- tokenizer.json +0 -0
- vocab.json +0 -0
README.md
CHANGED
@@ -13,14 +13,14 @@ widget:
|
|
13 |
- "Harry leapt forward, dodging Fingolfin's wildfire and reaching for the ring."
|
14 |
---
|
15 |
|
16 |
-
GPT-2 (including tokenizer) trained **from scratch** on some of my favorite books (about
|
17 |
|
18 |
-
It's only trained on an RTX 3090 for
|
19 |
|
20 |
- peak lr: 4e-4
|
21 |
- global batch size: 32
|
22 |
- weight decay: 0.01
|
23 |
-
- training steps:
|
24 |
- warmup steps: 1k
|
25 |
- lr decay: cosine
|
26 |
|
@@ -42,14 +42,20 @@ print(tokenizer.decode(output[0]))
|
|
42 |
|
43 |
Sample output:
|
44 |
|
45 |
-
> Daenerys kissed Gandalf, as the witcher hacked off Lord Voldemort's head with a brutal swing of Longclaw.
|
46 |
>
|
47 |
-
>
|
48 |
>
|
49 |
-
>
|
50 |
>
|
51 |
-
>
|
52 |
>
|
53 |
-
>
|
54 |
>
|
55 |
-
>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
13 |
- "Harry leapt forward, dodging Fingolfin's wildfire and reaching for the ring."
|
14 |
---
|
15 |
|
16 |
+
GPT-2 (including tokenizer) trained **from scratch** on some of my favorite books (about 31M words in total).
|
17 |
|
18 |
+
It's only trained on an RTX 3090 for three hours, so don't take it seriously, just have fun!
|
19 |
|
20 |
- peak lr: 4e-4
|
21 |
- global batch size: 32
|
22 |
- weight decay: 0.01
|
23 |
+
- training steps: 25k
|
24 |
- warmup steps: 1k
|
25 |
- lr decay: cosine
|
26 |
|
|
|
42 |
|
43 |
Sample output:
|
44 |
|
45 |
+
> Daenerys kissed Gandalf, as the witcher hacked off Lord Voldemort's head with a brutal swing of Longclaw. “Do you know what a warrior like that will do?”
|
46 |
>
|
47 |
+
> “I am a knight of seven,” Geralt said. “And how is this knight?”
|
48 |
>
|
49 |
+
> “Prince, it is known,” replied the witcher, “I am a knight of Solamnia, not a Knight. A knight of Solamnia in the name of Reorx, with elven armies at the head of his knights and knights. You do not even remember my name, elf. It was a good call.”
|
50 |
>
|
51 |
+
> “That name,” said the witcher, “what does it mean, elf?”
|
52 |
>
|
53 |
+
> “Some story,” said the old knight. “A good story from the Cataclysm.”
|
54 |
>
|
55 |
+
> The witcher snorted and looked at the witcher. “That is how you feel, elf. You don't understand why the knights won't be executed for them.”
|
56 |
+
>
|
57 |
+
> “That's because the knight is a king,” Chireadan said finally, “the king's elder son. Because he is, the knights would also be honorable with the title of knighthood. But a warrior would be better suited to have the honor to win.”
|
58 |
+
>
|
59 |
+
> “But…”
|
60 |
+
>
|
61 |
+
> “It's true,” said the old knight, “that, isn't entirely what you want. The lance of your sword is one of the best
|
config.json
CHANGED
@@ -31,7 +31,7 @@
|
|
31 |
"do_sample": true,
|
32 |
"max_length": 256,
|
33 |
"temperature": 1,
|
34 |
-
"top_p": 0.
|
35 |
}
|
36 |
},
|
37 |
"torch_dtype": "float32",
|
|
|
31 |
"do_sample": true,
|
32 |
"max_length": 256,
|
33 |
"temperature": 1,
|
34 |
+
"top_p": 0.9
|
35 |
}
|
36 |
},
|
37 |
"torch_dtype": "float32",
|
merges.txt
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 510395581
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6c9a1f8ac848d03d09eca02882e8ef27faa7d7995260eafab1fe5bc94712b406
|
3 |
size 510395581
|
tokenizer.json
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
vocab.json
CHANGED
The diff for this file is too large to render.
See raw diff
|
|