Training in progress, step 100
Browse files
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "/mnt/bn/xuruijie-llm/checkpoints/new_world/v1-distill-
|
3 |
"architectures": [
|
4 |
"MistralForCausalLM"
|
5 |
],
|
@@ -20,7 +20,7 @@
|
|
20 |
"sliding_window": 4096,
|
21 |
"tie_word_embeddings": false,
|
22 |
"torch_dtype": "bfloat16",
|
23 |
-
"transformers_version": "4.
|
24 |
"use_cache": false,
|
25 |
"vocab_size": 32002
|
26 |
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "/mnt/bn/xuruijie-llm/checkpoints/new_world/v1-distill-round2-0.5-diff-lr58",
|
3 |
"architectures": [
|
4 |
"MistralForCausalLM"
|
5 |
],
|
|
|
20 |
"sliding_window": 4096,
|
21 |
"tie_word_embeddings": false,
|
22 |
"torch_dtype": "bfloat16",
|
23 |
+
"transformers_version": "4.38.2",
|
24 |
"use_cache": false,
|
25 |
"vocab_size": 32002
|
26 |
}
|
runs/Jun04_11-43-48_n136-100-194/events.out.tfevents.1717472990.n136-100-194.3983374.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5941f4750df327c11d9045a1fe54b137808102976abcc232e5363bdd3392ede2
|
3 |
+
size 13101
|
tokenizer.json
CHANGED
@@ -152,7 +152,6 @@
|
|
152 |
"end_of_word_suffix": null,
|
153 |
"fuse_unk": true,
|
154 |
"byte_fallback": true,
|
155 |
-
"ignore_merges": false,
|
156 |
"vocab": {
|
157 |
"<unk>": 0,
|
158 |
"<s>": 1,
|
|
|
152 |
"end_of_word_suffix": null,
|
153 |
"fuse_unk": true,
|
154 |
"byte_fallback": true,
|
|
|
155 |
"vocab": {
|
156 |
"<unk>": 0,
|
157 |
"<s>": 1,
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:55a5fe1d8e345fca3fd945298a735f70df317cac7a0afe275b3577c9f2008a69
|
3 |
+
size 6264
|