Training in progress, step 500
Browse files- config.json +29 -0
- runs/Feb21_16-09-20_pop-os/events.out.tfevents.1708556965.pop-os.1697876.0 +3 -0
- runs/Feb21_16-12-24_pop-os/events.out.tfevents.1708557149.pop-os.1698373.0 +3 -0
- runs/Feb21_16-13-42_pop-os/events.out.tfevents.1708557227.pop-os.1698784.0 +3 -0
- runs/Feb21_16-20-37_pop-os/events.out.tfevents.1708557642.pop-os.1699149.0 +3 -0
- runs/Feb21_16-26-34_pop-os/events.out.tfevents.1708557999.pop-os.1699855.0 +3 -0
- runs/Feb21_16-29-21_pop-os/events.out.tfevents.1708558166.pop-os.1700185.0 +3 -0
- runs/Feb21_16-31-54_pop-os/events.out.tfevents.1708558319.pop-os.1700439.0 +3 -0
- runs/Feb21_16-36-04_pop-os/events.out.tfevents.1708558568.pop-os.1700783.0 +3 -0
- runs/Feb21_16-38-38_pop-os/events.out.tfevents.1708558723.pop-os.1701127.0 +3 -0
- runs/Feb21_16-43-14_pop-os/events.out.tfevents.1708558998.pop-os.1701791.0 +3 -0
- runs/Feb21_16-48-43_pop-os/events.out.tfevents.1708559328.pop-os.1702148.0 +3 -0
- runs/Feb21_16-51-30_pop-os/events.out.tfevents.1708559495.pop-os.1702415.0 +3 -0
- training_args.bin +3 -0
config.json
ADDED
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "TinyPixel/Llama-2-7B-bf16-sharded",
|
3 |
+
"architectures": [
|
4 |
+
"LlamaForCausalLM"
|
5 |
+
],
|
6 |
+
"attention_bias": false,
|
7 |
+
"attention_dropout": 0.0,
|
8 |
+
"bos_token_id": 1,
|
9 |
+
"eos_token_id": 2,
|
10 |
+
"hidden_act": "silu",
|
11 |
+
"hidden_size": 4096,
|
12 |
+
"initializer_range": 0.02,
|
13 |
+
"intermediate_size": 11008,
|
14 |
+
"max_position_embeddings": 2048,
|
15 |
+
"model_type": "llama",
|
16 |
+
"num_attention_heads": 32,
|
17 |
+
"num_hidden_layers": 32,
|
18 |
+
"num_key_value_heads": 32,
|
19 |
+
"pad_token_id": 0,
|
20 |
+
"pretraining_tp": 1,
|
21 |
+
"rms_norm_eps": 1e-05,
|
22 |
+
"rope_scaling": null,
|
23 |
+
"rope_theta": 10000.0,
|
24 |
+
"tie_word_embeddings": false,
|
25 |
+
"torch_dtype": "float32",
|
26 |
+
"transformers_version": "4.38.0",
|
27 |
+
"use_cache": true,
|
28 |
+
"vocab_size": 32000
|
29 |
+
}
|
runs/Feb21_16-09-20_pop-os/events.out.tfevents.1708556965.pop-os.1697876.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:480a2fab7c08c2f71c41c6f69dc4a16bf45fe9a468bc05603065c5fe553c04d7
|
3 |
+
size 4645
|
runs/Feb21_16-12-24_pop-os/events.out.tfevents.1708557149.pop-os.1698373.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a5c76baab2506b4c1dfa5ba2d68d3b2a23892eddcd8b619b4ec6d3f165218d34
|
3 |
+
size 4645
|
runs/Feb21_16-13-42_pop-os/events.out.tfevents.1708557227.pop-os.1698784.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4f1ba51b1865671883dab0ca567282a58f5a77d4d1642061c3e7e5040ccb7f5a
|
3 |
+
size 4645
|
runs/Feb21_16-20-37_pop-os/events.out.tfevents.1708557642.pop-os.1699149.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9e08299f5f18299af3144443548f8022e8828a176ba8353005267480f1d08f2b
|
3 |
+
size 4645
|
runs/Feb21_16-26-34_pop-os/events.out.tfevents.1708557999.pop-os.1699855.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:06d24d7a34d731c5a87a381a5faa379dfdbc0809b936f340d2bfc23a1e063cee
|
3 |
+
size 4645
|
runs/Feb21_16-29-21_pop-os/events.out.tfevents.1708558166.pop-os.1700185.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:888952f014ab9e8da6605274d5132e2273da98b8efb421d6e57046b9913177f3
|
3 |
+
size 4645
|
runs/Feb21_16-31-54_pop-os/events.out.tfevents.1708558319.pop-os.1700439.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f0898ce967d0757b099d7ba61c99d58c2e93c8ba1e713303f5a6ebcf907fa770
|
3 |
+
size 4645
|
runs/Feb21_16-36-04_pop-os/events.out.tfevents.1708558568.pop-os.1700783.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:41b8ca94ce095c8f4a3754fd068c6d9243b589c7793bfd53f1dabca9fa243ba3
|
3 |
+
size 4645
|
runs/Feb21_16-38-38_pop-os/events.out.tfevents.1708558723.pop-os.1701127.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ed8acd3438d5703e875fb897b012be12a55857546a4bb886b32a833843169b1e
|
3 |
+
size 4645
|
runs/Feb21_16-43-14_pop-os/events.out.tfevents.1708558998.pop-os.1701791.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6a60e2c2d8fdf9a7e3c2925606821797d9b019293a4f83c31c3318f608572e06
|
3 |
+
size 4645
|
runs/Feb21_16-48-43_pop-os/events.out.tfevents.1708559328.pop-os.1702148.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:78b5c89548380f56527d0e65772de48d2a1e84e1e0eefa534c0700e42c319310
|
3 |
+
size 4645
|
runs/Feb21_16-51-30_pop-os/events.out.tfevents.1708559495.pop-os.1702415.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5a78a9ba87fb0fb02addde3dc68f0b3226e8a68206d0f1389713606717ad0573
|
3 |
+
size 4856
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:73645fb58295d8d12765309234fe74726c344dfe2f5daae58e2bd81ab440e320
|
3 |
+
size 4984
|