Jon
commited on
Commit
·
eb03016
1
Parent(s):
be7ebfc
Upload 8 files
Browse files- README.md +32 -1
- adapter_config.json +21 -0
- adapter_model.bin +3 -0
- training_graph.json +1268 -0
- training_graph.png +0 -0
- training_log.json +18 -0
- training_parameters.json +37 -0
- training_prompt.json +3 -0
README.md
CHANGED
@@ -1,3 +1,34 @@
|
|
1 |
---
|
2 |
-
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
library_name: peft
|
3 |
---
|
4 |
+
## Training procedure
|
5 |
+
|
6 |
+
|
7 |
+
The following `bitsandbytes` quantization config was used during training:
|
8 |
+
- quant_method: QuantizationMethod.BITS_AND_BYTES
|
9 |
+
- load_in_8bit: False
|
10 |
+
- load_in_4bit: True
|
11 |
+
- llm_int8_threshold: 6.0
|
12 |
+
- llm_int8_skip_modules: None
|
13 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
14 |
+
- llm_int8_has_fp16_weight: False
|
15 |
+
- bnb_4bit_quant_type: nf4
|
16 |
+
- bnb_4bit_use_double_quant: True
|
17 |
+
- bnb_4bit_compute_dtype: float16
|
18 |
+
|
19 |
+
The following `bitsandbytes` quantization config was used during training:
|
20 |
+
- quant_method: QuantizationMethod.BITS_AND_BYTES
|
21 |
+
- load_in_8bit: False
|
22 |
+
- load_in_4bit: True
|
23 |
+
- llm_int8_threshold: 6.0
|
24 |
+
- llm_int8_skip_modules: None
|
25 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
26 |
+
- llm_int8_has_fp16_weight: False
|
27 |
+
- bnb_4bit_quant_type: nf4
|
28 |
+
- bnb_4bit_use_double_quant: True
|
29 |
+
- bnb_4bit_compute_dtype: float16
|
30 |
+
### Framework versions
|
31 |
+
|
32 |
+
- PEFT 0.5.0
|
33 |
+
|
34 |
+
- PEFT 0.5.0
|
adapter_config.json
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "models\\HuggingFaceH4_zephyr-7b-alpha",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": false,
|
6 |
+
"inference_mode": true,
|
7 |
+
"init_lora_weights": true,
|
8 |
+
"layers_pattern": null,
|
9 |
+
"layers_to_transform": null,
|
10 |
+
"lora_alpha": 64,
|
11 |
+
"lora_dropout": 0.05,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 32,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"q_proj",
|
18 |
+
"v_proj"
|
19 |
+
],
|
20 |
+
"task_type": "CAUSAL_LM"
|
21 |
+
}
|
adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:53e37f76541a681d7adc309759a6fec09199cefa4e57e670fb85c99c07e72fba
|
3 |
+
size 54572362
|
training_graph.json
ADDED
@@ -0,0 +1,1268 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"current_steps": 0,
|
4 |
+
"loss": 3.6926,
|
5 |
+
"learning_rate": 0.0,
|
6 |
+
"epoch": 0.01
|
7 |
+
},
|
8 |
+
{
|
9 |
+
"current_steps": 1,
|
10 |
+
"loss": 3.0107,
|
11 |
+
"learning_rate": 0.0,
|
12 |
+
"epoch": 0.03
|
13 |
+
},
|
14 |
+
{
|
15 |
+
"current_steps": 2,
|
16 |
+
"loss": 3.0688,
|
17 |
+
"learning_rate": 0.0,
|
18 |
+
"epoch": 0.04
|
19 |
+
},
|
20 |
+
{
|
21 |
+
"current_steps": 3,
|
22 |
+
"loss": 4.0176,
|
23 |
+
"learning_rate": 0.0,
|
24 |
+
"epoch": 0.06
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"current_steps": 4,
|
28 |
+
"loss": 4.0609,
|
29 |
+
"learning_rate": 0.0,
|
30 |
+
"epoch": 0.07
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"current_steps": 5,
|
34 |
+
"loss": 3.754,
|
35 |
+
"learning_rate": 0.0,
|
36 |
+
"epoch": 0.09
|
37 |
+
},
|
38 |
+
{
|
39 |
+
"current_steps": 6,
|
40 |
+
"loss": 3.1981,
|
41 |
+
"learning_rate": 0.0,
|
42 |
+
"epoch": 0.1
|
43 |
+
},
|
44 |
+
{
|
45 |
+
"current_steps": 7,
|
46 |
+
"loss": 2.7138,
|
47 |
+
"learning_rate": 0.0,
|
48 |
+
"epoch": 0.11
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"current_steps": 8,
|
52 |
+
"loss": 3.8803,
|
53 |
+
"learning_rate": 0.0,
|
54 |
+
"epoch": 0.13
|
55 |
+
},
|
56 |
+
{
|
57 |
+
"current_steps": 9,
|
58 |
+
"loss": 3.5793,
|
59 |
+
"learning_rate": 2.9999999999999997e-06,
|
60 |
+
"epoch": 0.14
|
61 |
+
},
|
62 |
+
{
|
63 |
+
"current_steps": 10,
|
64 |
+
"loss": 2.8589,
|
65 |
+
"learning_rate": 5.999999999999999e-06,
|
66 |
+
"epoch": 0.16
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"current_steps": 11,
|
70 |
+
"loss": 2.6536,
|
71 |
+
"learning_rate": 8.999999999999999e-06,
|
72 |
+
"epoch": 0.17
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"current_steps": 12,
|
76 |
+
"loss": 4.1845,
|
77 |
+
"learning_rate": 1.1999999999999999e-05,
|
78 |
+
"epoch": 0.19
|
79 |
+
},
|
80 |
+
{
|
81 |
+
"current_steps": 13,
|
82 |
+
"loss": 3.7447,
|
83 |
+
"learning_rate": 1.4999999999999999e-05,
|
84 |
+
"epoch": 0.2
|
85 |
+
},
|
86 |
+
{
|
87 |
+
"current_steps": 14,
|
88 |
+
"loss": 2.9558,
|
89 |
+
"learning_rate": 1.7999999999999997e-05,
|
90 |
+
"epoch": 0.21
|
91 |
+
},
|
92 |
+
{
|
93 |
+
"current_steps": 15,
|
94 |
+
"loss": 3.7552,
|
95 |
+
"learning_rate": 2.1e-05,
|
96 |
+
"epoch": 0.23
|
97 |
+
},
|
98 |
+
{
|
99 |
+
"current_steps": 16,
|
100 |
+
"loss": 3.7336,
|
101 |
+
"learning_rate": 2.3999999999999997e-05,
|
102 |
+
"epoch": 0.24
|
103 |
+
},
|
104 |
+
{
|
105 |
+
"current_steps": 17,
|
106 |
+
"loss": 2.484,
|
107 |
+
"learning_rate": 2.6999999999999996e-05,
|
108 |
+
"epoch": 0.26
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"current_steps": 18,
|
112 |
+
"loss": 3.7203,
|
113 |
+
"learning_rate": 2.9999999999999997e-05,
|
114 |
+
"epoch": 0.27
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"current_steps": 19,
|
118 |
+
"loss": 3.6129,
|
119 |
+
"learning_rate": 2.9999999999999997e-05,
|
120 |
+
"epoch": 0.29
|
121 |
+
},
|
122 |
+
{
|
123 |
+
"current_steps": 20,
|
124 |
+
"loss": 2.2497,
|
125 |
+
"learning_rate": 3.2999999999999996e-05,
|
126 |
+
"epoch": 0.3
|
127 |
+
},
|
128 |
+
{
|
129 |
+
"current_steps": 21,
|
130 |
+
"loss": 3.2732,
|
131 |
+
"learning_rate": 3.5999999999999994e-05,
|
132 |
+
"epoch": 0.31
|
133 |
+
},
|
134 |
+
{
|
135 |
+
"current_steps": 22,
|
136 |
+
"loss": 2.4979,
|
137 |
+
"learning_rate": 3.9e-05,
|
138 |
+
"epoch": 0.33
|
139 |
+
},
|
140 |
+
{
|
141 |
+
"current_steps": 23,
|
142 |
+
"loss": 3.043,
|
143 |
+
"learning_rate": 4.2e-05,
|
144 |
+
"epoch": 0.34
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"current_steps": 24,
|
148 |
+
"loss": 3.2435,
|
149 |
+
"learning_rate": 4.4999999999999996e-05,
|
150 |
+
"epoch": 0.36
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"current_steps": 25,
|
154 |
+
"loss": 3.3493,
|
155 |
+
"learning_rate": 4.7999999999999994e-05,
|
156 |
+
"epoch": 0.37
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"current_steps": 26,
|
160 |
+
"loss": 2.6339,
|
161 |
+
"learning_rate": 5.1e-05,
|
162 |
+
"epoch": 0.39
|
163 |
+
},
|
164 |
+
{
|
165 |
+
"current_steps": 27,
|
166 |
+
"loss": 2.9486,
|
167 |
+
"learning_rate": 5.399999999999999e-05,
|
168 |
+
"epoch": 0.4
|
169 |
+
},
|
170 |
+
{
|
171 |
+
"current_steps": 28,
|
172 |
+
"loss": 2.3865,
|
173 |
+
"learning_rate": 5.6999999999999996e-05,
|
174 |
+
"epoch": 0.41
|
175 |
+
},
|
176 |
+
{
|
177 |
+
"current_steps": 29,
|
178 |
+
"loss": 2.3687,
|
179 |
+
"learning_rate": 5.9999999999999995e-05,
|
180 |
+
"epoch": 0.43
|
181 |
+
},
|
182 |
+
{
|
183 |
+
"current_steps": 30,
|
184 |
+
"loss": 2.3417,
|
185 |
+
"learning_rate": 6.299999999999999e-05,
|
186 |
+
"epoch": 0.44
|
187 |
+
},
|
188 |
+
{
|
189 |
+
"current_steps": 31,
|
190 |
+
"loss": 2.5443,
|
191 |
+
"learning_rate": 6.599999999999999e-05,
|
192 |
+
"epoch": 0.46
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"current_steps": 32,
|
196 |
+
"loss": 2.4717,
|
197 |
+
"learning_rate": 6.9e-05,
|
198 |
+
"epoch": 0.47
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"current_steps": 33,
|
202 |
+
"loss": 2.2291,
|
203 |
+
"learning_rate": 7.199999999999999e-05,
|
204 |
+
"epoch": 0.49
|
205 |
+
},
|
206 |
+
{
|
207 |
+
"current_steps": 34,
|
208 |
+
"loss": 2.4815,
|
209 |
+
"learning_rate": 7.5e-05,
|
210 |
+
"epoch": 0.5
|
211 |
+
},
|
212 |
+
{
|
213 |
+
"current_steps": 35,
|
214 |
+
"loss": 2.4468,
|
215 |
+
"learning_rate": 7.8e-05,
|
216 |
+
"epoch": 0.51
|
217 |
+
},
|
218 |
+
{
|
219 |
+
"current_steps": 36,
|
220 |
+
"loss": 2.4246,
|
221 |
+
"learning_rate": 8.1e-05,
|
222 |
+
"epoch": 0.53
|
223 |
+
},
|
224 |
+
{
|
225 |
+
"current_steps": 37,
|
226 |
+
"loss": 2.3154,
|
227 |
+
"learning_rate": 8.4e-05,
|
228 |
+
"epoch": 0.54
|
229 |
+
},
|
230 |
+
{
|
231 |
+
"current_steps": 38,
|
232 |
+
"loss": 1.9939,
|
233 |
+
"learning_rate": 8.699999999999999e-05,
|
234 |
+
"epoch": 0.56
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"current_steps": 39,
|
238 |
+
"loss": 1.8099,
|
239 |
+
"learning_rate": 8.999999999999999e-05,
|
240 |
+
"epoch": 0.57
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"current_steps": 40,
|
244 |
+
"loss": 1.865,
|
245 |
+
"learning_rate": 9.3e-05,
|
246 |
+
"epoch": 0.59
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"current_steps": 41,
|
250 |
+
"loss": 1.4322,
|
251 |
+
"learning_rate": 9.599999999999999e-05,
|
252 |
+
"epoch": 0.6
|
253 |
+
},
|
254 |
+
{
|
255 |
+
"current_steps": 42,
|
256 |
+
"loss": 2.3252,
|
257 |
+
"learning_rate": 9.9e-05,
|
258 |
+
"epoch": 0.61
|
259 |
+
},
|
260 |
+
{
|
261 |
+
"current_steps": 43,
|
262 |
+
"loss": 1.7521,
|
263 |
+
"learning_rate": 0.000102,
|
264 |
+
"epoch": 0.63
|
265 |
+
},
|
266 |
+
{
|
267 |
+
"current_steps": 44,
|
268 |
+
"loss": 2.145,
|
269 |
+
"learning_rate": 0.00010499999999999999,
|
270 |
+
"epoch": 0.64
|
271 |
+
},
|
272 |
+
{
|
273 |
+
"current_steps": 45,
|
274 |
+
"loss": 1.6641,
|
275 |
+
"learning_rate": 0.00010799999999999998,
|
276 |
+
"epoch": 0.66
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"current_steps": 46,
|
280 |
+
"loss": 1.8938,
|
281 |
+
"learning_rate": 0.00011099999999999999,
|
282 |
+
"epoch": 0.67
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"current_steps": 47,
|
286 |
+
"loss": 1.5194,
|
287 |
+
"learning_rate": 0.00011399999999999999,
|
288 |
+
"epoch": 0.69
|
289 |
+
},
|
290 |
+
{
|
291 |
+
"current_steps": 48,
|
292 |
+
"loss": 1.4478,
|
293 |
+
"learning_rate": 0.000117,
|
294 |
+
"epoch": 0.7
|
295 |
+
},
|
296 |
+
{
|
297 |
+
"current_steps": 49,
|
298 |
+
"loss": 1.9414,
|
299 |
+
"learning_rate": 0.00011999999999999999,
|
300 |
+
"epoch": 0.71
|
301 |
+
},
|
302 |
+
{
|
303 |
+
"current_steps": 50,
|
304 |
+
"loss": 1.5601,
|
305 |
+
"learning_rate": 0.00012299999999999998,
|
306 |
+
"epoch": 0.73
|
307 |
+
},
|
308 |
+
{
|
309 |
+
"current_steps": 51,
|
310 |
+
"loss": 1.9015,
|
311 |
+
"learning_rate": 0.00012599999999999997,
|
312 |
+
"epoch": 0.74
|
313 |
+
},
|
314 |
+
{
|
315 |
+
"current_steps": 52,
|
316 |
+
"loss": 1.7384,
|
317 |
+
"learning_rate": 0.000129,
|
318 |
+
"epoch": 0.76
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"current_steps": 53,
|
322 |
+
"loss": 1.7282,
|
323 |
+
"learning_rate": 0.00013199999999999998,
|
324 |
+
"epoch": 0.77
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"current_steps": 54,
|
328 |
+
"loss": 1.6877,
|
329 |
+
"learning_rate": 0.000135,
|
330 |
+
"epoch": 0.79
|
331 |
+
},
|
332 |
+
{
|
333 |
+
"current_steps": 55,
|
334 |
+
"loss": 1.0647,
|
335 |
+
"learning_rate": 0.000138,
|
336 |
+
"epoch": 0.8
|
337 |
+
},
|
338 |
+
{
|
339 |
+
"current_steps": 56,
|
340 |
+
"loss": 1.755,
|
341 |
+
"learning_rate": 0.00014099999999999998,
|
342 |
+
"epoch": 0.81
|
343 |
+
},
|
344 |
+
{
|
345 |
+
"current_steps": 57,
|
346 |
+
"loss": 1.516,
|
347 |
+
"learning_rate": 0.00014399999999999998,
|
348 |
+
"epoch": 0.83
|
349 |
+
},
|
350 |
+
{
|
351 |
+
"current_steps": 58,
|
352 |
+
"loss": 1.3498,
|
353 |
+
"learning_rate": 0.000147,
|
354 |
+
"epoch": 0.84
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"current_steps": 59,
|
358 |
+
"loss": 1.6789,
|
359 |
+
"learning_rate": 0.00015,
|
360 |
+
"epoch": 0.86
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"current_steps": 60,
|
364 |
+
"loss": 1.3651,
|
365 |
+
"learning_rate": 0.00015299999999999998,
|
366 |
+
"epoch": 0.87
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"current_steps": 61,
|
370 |
+
"loss": 1.2967,
|
371 |
+
"learning_rate": 0.000156,
|
372 |
+
"epoch": 0.89
|
373 |
+
},
|
374 |
+
{
|
375 |
+
"current_steps": 62,
|
376 |
+
"loss": 1.3998,
|
377 |
+
"learning_rate": 0.000159,
|
378 |
+
"epoch": 0.9
|
379 |
+
},
|
380 |
+
{
|
381 |
+
"current_steps": 63,
|
382 |
+
"loss": 1.1989,
|
383 |
+
"learning_rate": 0.000162,
|
384 |
+
"epoch": 0.91
|
385 |
+
},
|
386 |
+
{
|
387 |
+
"current_steps": 64,
|
388 |
+
"loss": 1.029,
|
389 |
+
"learning_rate": 0.000165,
|
390 |
+
"epoch": 0.93
|
391 |
+
},
|
392 |
+
{
|
393 |
+
"current_steps": 65,
|
394 |
+
"loss": 1.4441,
|
395 |
+
"learning_rate": 0.000168,
|
396 |
+
"epoch": 0.94
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"current_steps": 66,
|
400 |
+
"loss": 1.1804,
|
401 |
+
"learning_rate": 0.00017099999999999998,
|
402 |
+
"epoch": 0.96
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"current_steps": 67,
|
406 |
+
"loss": 1.2868,
|
407 |
+
"learning_rate": 0.00017399999999999997,
|
408 |
+
"epoch": 0.97
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"current_steps": 68,
|
412 |
+
"loss": 1.1724,
|
413 |
+
"learning_rate": 0.00017699999999999997,
|
414 |
+
"epoch": 0.99
|
415 |
+
},
|
416 |
+
{
|
417 |
+
"current_steps": 69,
|
418 |
+
"loss": 1.4346,
|
419 |
+
"learning_rate": 0.00017999999999999998,
|
420 |
+
"epoch": 1.0
|
421 |
+
},
|
422 |
+
{
|
423 |
+
"current_steps": 70,
|
424 |
+
"loss": 1.0183,
|
425 |
+
"learning_rate": 0.00018299999999999998,
|
426 |
+
"epoch": 1.01
|
427 |
+
},
|
428 |
+
{
|
429 |
+
"current_steps": 71,
|
430 |
+
"loss": 1.1259,
|
431 |
+
"learning_rate": 0.000186,
|
432 |
+
"epoch": 1.03
|
433 |
+
},
|
434 |
+
{
|
435 |
+
"current_steps": 72,
|
436 |
+
"loss": 1.1713,
|
437 |
+
"learning_rate": 0.00018899999999999999,
|
438 |
+
"epoch": 1.04
|
439 |
+
},
|
440 |
+
{
|
441 |
+
"current_steps": 73,
|
442 |
+
"loss": 1.0773,
|
443 |
+
"learning_rate": 0.00019199999999999998,
|
444 |
+
"epoch": 1.06
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"current_steps": 74,
|
448 |
+
"loss": 1.2956,
|
449 |
+
"learning_rate": 0.000195,
|
450 |
+
"epoch": 1.07
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"current_steps": 75,
|
454 |
+
"loss": 0.9359,
|
455 |
+
"learning_rate": 0.000198,
|
456 |
+
"epoch": 1.09
|
457 |
+
},
|
458 |
+
{
|
459 |
+
"current_steps": 76,
|
460 |
+
"loss": 0.9838,
|
461 |
+
"learning_rate": 0.000201,
|
462 |
+
"epoch": 1.1
|
463 |
+
},
|
464 |
+
{
|
465 |
+
"current_steps": 77,
|
466 |
+
"loss": 0.9488,
|
467 |
+
"learning_rate": 0.000204,
|
468 |
+
"epoch": 1.11
|
469 |
+
},
|
470 |
+
{
|
471 |
+
"current_steps": 78,
|
472 |
+
"loss": 1.0748,
|
473 |
+
"learning_rate": 0.00020699999999999996,
|
474 |
+
"epoch": 1.13
|
475 |
+
},
|
476 |
+
{
|
477 |
+
"current_steps": 79,
|
478 |
+
"loss": 1.0914,
|
479 |
+
"learning_rate": 0.00020999999999999998,
|
480 |
+
"epoch": 1.14
|
481 |
+
},
|
482 |
+
{
|
483 |
+
"current_steps": 80,
|
484 |
+
"loss": 0.5364,
|
485 |
+
"learning_rate": 0.00021299999999999997,
|
486 |
+
"epoch": 1.16
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"current_steps": 81,
|
490 |
+
"loss": 0.4909,
|
491 |
+
"learning_rate": 0.00021599999999999996,
|
492 |
+
"epoch": 1.17
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"current_steps": 82,
|
496 |
+
"loss": 1.0176,
|
497 |
+
"learning_rate": 0.00021899999999999998,
|
498 |
+
"epoch": 1.19
|
499 |
+
},
|
500 |
+
{
|
501 |
+
"current_steps": 83,
|
502 |
+
"loss": 1.0543,
|
503 |
+
"learning_rate": 0.00022199999999999998,
|
504 |
+
"epoch": 1.2
|
505 |
+
},
|
506 |
+
{
|
507 |
+
"current_steps": 84,
|
508 |
+
"loss": 0.9092,
|
509 |
+
"learning_rate": 0.000225,
|
510 |
+
"epoch": 1.21
|
511 |
+
},
|
512 |
+
{
|
513 |
+
"current_steps": 85,
|
514 |
+
"loss": 0.5678,
|
515 |
+
"learning_rate": 0.00022799999999999999,
|
516 |
+
"epoch": 1.23
|
517 |
+
},
|
518 |
+
{
|
519 |
+
"current_steps": 86,
|
520 |
+
"loss": 0.8385,
|
521 |
+
"learning_rate": 0.00023099999999999998,
|
522 |
+
"epoch": 1.24
|
523 |
+
},
|
524 |
+
{
|
525 |
+
"current_steps": 87,
|
526 |
+
"loss": 1.6241,
|
527 |
+
"learning_rate": 0.000234,
|
528 |
+
"epoch": 1.26
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"current_steps": 88,
|
532 |
+
"loss": 0.8294,
|
533 |
+
"learning_rate": 0.000237,
|
534 |
+
"epoch": 1.27
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"current_steps": 89,
|
538 |
+
"loss": 1.2953,
|
539 |
+
"learning_rate": 0.00023999999999999998,
|
540 |
+
"epoch": 1.29
|
541 |
+
},
|
542 |
+
{
|
543 |
+
"current_steps": 90,
|
544 |
+
"loss": 0.5297,
|
545 |
+
"learning_rate": 0.000243,
|
546 |
+
"epoch": 1.3
|
547 |
+
},
|
548 |
+
{
|
549 |
+
"current_steps": 91,
|
550 |
+
"loss": 1.1592,
|
551 |
+
"learning_rate": 0.00024599999999999996,
|
552 |
+
"epoch": 1.31
|
553 |
+
},
|
554 |
+
{
|
555 |
+
"current_steps": 92,
|
556 |
+
"loss": 1.1055,
|
557 |
+
"learning_rate": 0.000249,
|
558 |
+
"epoch": 1.33
|
559 |
+
},
|
560 |
+
{
|
561 |
+
"current_steps": 93,
|
562 |
+
"loss": 0.9804,
|
563 |
+
"learning_rate": 0.00025199999999999995,
|
564 |
+
"epoch": 1.34
|
565 |
+
},
|
566 |
+
{
|
567 |
+
"current_steps": 94,
|
568 |
+
"loss": 0.8683,
|
569 |
+
"learning_rate": 0.00025499999999999996,
|
570 |
+
"epoch": 1.36
|
571 |
+
},
|
572 |
+
{
|
573 |
+
"current_steps": 95,
|
574 |
+
"loss": 0.9698,
|
575 |
+
"learning_rate": 0.000258,
|
576 |
+
"epoch": 1.37
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"current_steps": 96,
|
580 |
+
"loss": 1.037,
|
581 |
+
"learning_rate": 0.000261,
|
582 |
+
"epoch": 1.39
|
583 |
+
},
|
584 |
+
{
|
585 |
+
"current_steps": 97,
|
586 |
+
"loss": 0.5479,
|
587 |
+
"learning_rate": 0.00026399999999999997,
|
588 |
+
"epoch": 1.4
|
589 |
+
},
|
590 |
+
{
|
591 |
+
"current_steps": 98,
|
592 |
+
"loss": 0.598,
|
593 |
+
"learning_rate": 0.000267,
|
594 |
+
"epoch": 1.41
|
595 |
+
},
|
596 |
+
{
|
597 |
+
"current_steps": 99,
|
598 |
+
"loss": 0.3627,
|
599 |
+
"learning_rate": 0.00027,
|
600 |
+
"epoch": 1.43
|
601 |
+
},
|
602 |
+
{
|
603 |
+
"current_steps": 100,
|
604 |
+
"loss": 0.8042,
|
605 |
+
"learning_rate": 0.00027299999999999997,
|
606 |
+
"epoch": 1.44
|
607 |
+
},
|
608 |
+
{
|
609 |
+
"current_steps": 101,
|
610 |
+
"loss": 1.0378,
|
611 |
+
"learning_rate": 0.000276,
|
612 |
+
"epoch": 1.46
|
613 |
+
},
|
614 |
+
{
|
615 |
+
"current_steps": 102,
|
616 |
+
"loss": 1.0192,
|
617 |
+
"learning_rate": 0.000279,
|
618 |
+
"epoch": 1.47
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"current_steps": 103,
|
622 |
+
"loss": 1.4654,
|
623 |
+
"learning_rate": 0.00028199999999999997,
|
624 |
+
"epoch": 1.49
|
625 |
+
},
|
626 |
+
{
|
627 |
+
"current_steps": 104,
|
628 |
+
"loss": 1.0614,
|
629 |
+
"learning_rate": 0.000285,
|
630 |
+
"epoch": 1.5
|
631 |
+
},
|
632 |
+
{
|
633 |
+
"current_steps": 105,
|
634 |
+
"loss": 0.779,
|
635 |
+
"learning_rate": 0.00028799999999999995,
|
636 |
+
"epoch": 1.51
|
637 |
+
},
|
638 |
+
{
|
639 |
+
"current_steps": 106,
|
640 |
+
"loss": 1.0255,
|
641 |
+
"learning_rate": 0.00029099999999999997,
|
642 |
+
"epoch": 1.53
|
643 |
+
},
|
644 |
+
{
|
645 |
+
"current_steps": 107,
|
646 |
+
"loss": 1.0924,
|
647 |
+
"learning_rate": 0.000294,
|
648 |
+
"epoch": 1.54
|
649 |
+
},
|
650 |
+
{
|
651 |
+
"current_steps": 108,
|
652 |
+
"loss": 1.1016,
|
653 |
+
"learning_rate": 0.00029699999999999996,
|
654 |
+
"epoch": 1.56
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"current_steps": 109,
|
658 |
+
"loss": 0.8748,
|
659 |
+
"learning_rate": 0.0003,
|
660 |
+
"epoch": 1.57
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"current_steps": 110,
|
664 |
+
"loss": 0.8048,
|
665 |
+
"learning_rate": 0.00029727272727272724,
|
666 |
+
"epoch": 1.59
|
667 |
+
},
|
668 |
+
{
|
669 |
+
"current_steps": 111,
|
670 |
+
"loss": 1.0229,
|
671 |
+
"learning_rate": 0.0002945454545454545,
|
672 |
+
"epoch": 1.6
|
673 |
+
},
|
674 |
+
{
|
675 |
+
"current_steps": 112,
|
676 |
+
"loss": 0.9677,
|
677 |
+
"learning_rate": 0.0002918181818181818,
|
678 |
+
"epoch": 1.61
|
679 |
+
},
|
680 |
+
{
|
681 |
+
"current_steps": 113,
|
682 |
+
"loss": 0.5605,
|
683 |
+
"learning_rate": 0.00028909090909090904,
|
684 |
+
"epoch": 1.63
|
685 |
+
},
|
686 |
+
{
|
687 |
+
"current_steps": 114,
|
688 |
+
"loss": 0.9392,
|
689 |
+
"learning_rate": 0.00028636363636363636,
|
690 |
+
"epoch": 1.64
|
691 |
+
},
|
692 |
+
{
|
693 |
+
"current_steps": 115,
|
694 |
+
"loss": 1.0068,
|
695 |
+
"learning_rate": 0.0002836363636363636,
|
696 |
+
"epoch": 1.66
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"current_steps": 116,
|
700 |
+
"loss": 0.898,
|
701 |
+
"learning_rate": 0.0002809090909090909,
|
702 |
+
"epoch": 1.67
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"current_steps": 117,
|
706 |
+
"loss": 1.1297,
|
707 |
+
"learning_rate": 0.00027818181818181815,
|
708 |
+
"epoch": 1.69
|
709 |
+
},
|
710 |
+
{
|
711 |
+
"current_steps": 118,
|
712 |
+
"loss": 0.8696,
|
713 |
+
"learning_rate": 0.0002754545454545454,
|
714 |
+
"epoch": 1.7
|
715 |
+
},
|
716 |
+
{
|
717 |
+
"current_steps": 119,
|
718 |
+
"loss": 0.975,
|
719 |
+
"learning_rate": 0.0002727272727272727,
|
720 |
+
"epoch": 1.71
|
721 |
+
},
|
722 |
+
{
|
723 |
+
"current_steps": 120,
|
724 |
+
"loss": 1.107,
|
725 |
+
"learning_rate": 0.00027,
|
726 |
+
"epoch": 1.73
|
727 |
+
},
|
728 |
+
{
|
729 |
+
"current_steps": 121,
|
730 |
+
"loss": 1.1696,
|
731 |
+
"learning_rate": 0.0002672727272727272,
|
732 |
+
"epoch": 1.74
|
733 |
+
},
|
734 |
+
{
|
735 |
+
"current_steps": 122,
|
736 |
+
"loss": 0.7181,
|
737 |
+
"learning_rate": 0.00026454545454545453,
|
738 |
+
"epoch": 1.76
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"current_steps": 123,
|
742 |
+
"loss": 0.4008,
|
743 |
+
"learning_rate": 0.0002618181818181818,
|
744 |
+
"epoch": 1.77
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"current_steps": 124,
|
748 |
+
"loss": 0.6227,
|
749 |
+
"learning_rate": 0.00025909090909090907,
|
750 |
+
"epoch": 1.79
|
751 |
+
},
|
752 |
+
{
|
753 |
+
"current_steps": 125,
|
754 |
+
"loss": 1.4091,
|
755 |
+
"learning_rate": 0.00025636363636363633,
|
756 |
+
"epoch": 1.8
|
757 |
+
},
|
758 |
+
{
|
759 |
+
"current_steps": 126,
|
760 |
+
"loss": 0.7953,
|
761 |
+
"learning_rate": 0.0002536363636363636,
|
762 |
+
"epoch": 1.81
|
763 |
+
},
|
764 |
+
{
|
765 |
+
"current_steps": 127,
|
766 |
+
"loss": 1.1635,
|
767 |
+
"learning_rate": 0.00025090909090909086,
|
768 |
+
"epoch": 1.83
|
769 |
+
},
|
770 |
+
{
|
771 |
+
"current_steps": 128,
|
772 |
+
"loss": 0.6109,
|
773 |
+
"learning_rate": 0.0002481818181818182,
|
774 |
+
"epoch": 1.84
|
775 |
+
},
|
776 |
+
{
|
777 |
+
"current_steps": 129,
|
778 |
+
"loss": 0.7244,
|
779 |
+
"learning_rate": 0.00024545454545454545,
|
780 |
+
"epoch": 1.86
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"current_steps": 130,
|
784 |
+
"loss": 1.0868,
|
785 |
+
"learning_rate": 0.0002427272727272727,
|
786 |
+
"epoch": 1.87
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"current_steps": 131,
|
790 |
+
"loss": 1.1968,
|
791 |
+
"learning_rate": 0.00023999999999999998,
|
792 |
+
"epoch": 1.89
|
793 |
+
},
|
794 |
+
{
|
795 |
+
"current_steps": 132,
|
796 |
+
"loss": 1.109,
|
797 |
+
"learning_rate": 0.00023727272727272724,
|
798 |
+
"epoch": 1.9
|
799 |
+
},
|
800 |
+
{
|
801 |
+
"current_steps": 133,
|
802 |
+
"loss": 0.8744,
|
803 |
+
"learning_rate": 0.00023454545454545454,
|
804 |
+
"epoch": 1.91
|
805 |
+
},
|
806 |
+
{
|
807 |
+
"current_steps": 134,
|
808 |
+
"loss": 0.6971,
|
809 |
+
"learning_rate": 0.0002318181818181818,
|
810 |
+
"epoch": 1.93
|
811 |
+
},
|
812 |
+
{
|
813 |
+
"current_steps": 135,
|
814 |
+
"loss": 0.827,
|
815 |
+
"learning_rate": 0.00022909090909090907,
|
816 |
+
"epoch": 1.94
|
817 |
+
},
|
818 |
+
{
|
819 |
+
"current_steps": 136,
|
820 |
+
"loss": 0.8474,
|
821 |
+
"learning_rate": 0.00022636363636363633,
|
822 |
+
"epoch": 1.96
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"current_steps": 137,
|
826 |
+
"loss": 0.8051,
|
827 |
+
"learning_rate": 0.00022363636363636363,
|
828 |
+
"epoch": 1.97
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"current_steps": 138,
|
832 |
+
"loss": 0.8532,
|
833 |
+
"learning_rate": 0.0002209090909090909,
|
834 |
+
"epoch": 1.99
|
835 |
+
},
|
836 |
+
{
|
837 |
+
"current_steps": 139,
|
838 |
+
"loss": 0.8277,
|
839 |
+
"learning_rate": 0.00021818181818181816,
|
840 |
+
"epoch": 2.0
|
841 |
+
},
|
842 |
+
{
|
843 |
+
"current_steps": 140,
|
844 |
+
"loss": 0.3926,
|
845 |
+
"learning_rate": 0.00021545454545454542,
|
846 |
+
"epoch": 2.01
|
847 |
+
},
|
848 |
+
{
|
849 |
+
"current_steps": 141,
|
850 |
+
"loss": 0.5717,
|
851 |
+
"learning_rate": 0.00021272727272727272,
|
852 |
+
"epoch": 2.03
|
853 |
+
},
|
854 |
+
{
|
855 |
+
"current_steps": 142,
|
856 |
+
"loss": 0.6956,
|
857 |
+
"learning_rate": 0.00020999999999999998,
|
858 |
+
"epoch": 2.04
|
859 |
+
},
|
860 |
+
{
|
861 |
+
"current_steps": 143,
|
862 |
+
"loss": 0.6353,
|
863 |
+
"learning_rate": 0.00020727272727272725,
|
864 |
+
"epoch": 2.06
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"current_steps": 144,
|
868 |
+
"loss": 0.4248,
|
869 |
+
"learning_rate": 0.0002045454545454545,
|
870 |
+
"epoch": 2.07
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"current_steps": 145,
|
874 |
+
"loss": 0.6299,
|
875 |
+
"learning_rate": 0.0002018181818181818,
|
876 |
+
"epoch": 2.09
|
877 |
+
},
|
878 |
+
{
|
879 |
+
"current_steps": 146,
|
880 |
+
"loss": 0.3415,
|
881 |
+
"learning_rate": 0.0001990909090909091,
|
882 |
+
"epoch": 2.1
|
883 |
+
},
|
884 |
+
{
|
885 |
+
"current_steps": 147,
|
886 |
+
"loss": 0.4788,
|
887 |
+
"learning_rate": 0.00019636363636363634,
|
888 |
+
"epoch": 2.11
|
889 |
+
},
|
890 |
+
{
|
891 |
+
"current_steps": 148,
|
892 |
+
"loss": 0.6898,
|
893 |
+
"learning_rate": 0.00019363636363636363,
|
894 |
+
"epoch": 2.13
|
895 |
+
},
|
896 |
+
{
|
897 |
+
"current_steps": 149,
|
898 |
+
"loss": 0.4212,
|
899 |
+
"learning_rate": 0.0001909090909090909,
|
900 |
+
"epoch": 2.14
|
901 |
+
},
|
902 |
+
{
|
903 |
+
"current_steps": 150,
|
904 |
+
"loss": 0.6855,
|
905 |
+
"learning_rate": 0.0001881818181818182,
|
906 |
+
"epoch": 2.16
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"current_steps": 151,
|
910 |
+
"loss": 0.4411,
|
911 |
+
"learning_rate": 0.00018545454545454543,
|
912 |
+
"epoch": 2.17
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"current_steps": 152,
|
916 |
+
"loss": 0.4706,
|
917 |
+
"learning_rate": 0.00018272727272727272,
|
918 |
+
"epoch": 2.19
|
919 |
+
},
|
920 |
+
{
|
921 |
+
"current_steps": 153,
|
922 |
+
"loss": 0.6222,
|
923 |
+
"learning_rate": 0.00017999999999999998,
|
924 |
+
"epoch": 2.2
|
925 |
+
},
|
926 |
+
{
|
927 |
+
"current_steps": 154,
|
928 |
+
"loss": 0.6584,
|
929 |
+
"learning_rate": 0.00017727272727272728,
|
930 |
+
"epoch": 2.21
|
931 |
+
},
|
932 |
+
{
|
933 |
+
"current_steps": 155,
|
934 |
+
"loss": 0.6037,
|
935 |
+
"learning_rate": 0.00017454545454545452,
|
936 |
+
"epoch": 2.23
|
937 |
+
},
|
938 |
+
{
|
939 |
+
"current_steps": 156,
|
940 |
+
"loss": 0.6478,
|
941 |
+
"learning_rate": 0.0001718181818181818,
|
942 |
+
"epoch": 2.24
|
943 |
+
},
|
944 |
+
{
|
945 |
+
"current_steps": 157,
|
946 |
+
"loss": 0.4591,
|
947 |
+
"learning_rate": 0.00016909090909090907,
|
948 |
+
"epoch": 2.26
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"current_steps": 158,
|
952 |
+
"loss": 0.4085,
|
953 |
+
"learning_rate": 0.00016636363636363637,
|
954 |
+
"epoch": 2.27
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"current_steps": 159,
|
958 |
+
"loss": 0.6747,
|
959 |
+
"learning_rate": 0.0001636363636363636,
|
960 |
+
"epoch": 2.29
|
961 |
+
},
|
962 |
+
{
|
963 |
+
"current_steps": 160,
|
964 |
+
"loss": 0.4008,
|
965 |
+
"learning_rate": 0.0001609090909090909,
|
966 |
+
"epoch": 2.3
|
967 |
+
},
|
968 |
+
{
|
969 |
+
"current_steps": 161,
|
970 |
+
"loss": 0.671,
|
971 |
+
"learning_rate": 0.00015818181818181816,
|
972 |
+
"epoch": 2.31
|
973 |
+
},
|
974 |
+
{
|
975 |
+
"current_steps": 162,
|
976 |
+
"loss": 0.593,
|
977 |
+
"learning_rate": 0.00015545454545454546,
|
978 |
+
"epoch": 2.33
|
979 |
+
},
|
980 |
+
{
|
981 |
+
"current_steps": 163,
|
982 |
+
"loss": 0.4881,
|
983 |
+
"learning_rate": 0.0001527272727272727,
|
984 |
+
"epoch": 2.34
|
985 |
+
},
|
986 |
+
{
|
987 |
+
"current_steps": 164,
|
988 |
+
"loss": 0.6749,
|
989 |
+
"learning_rate": 0.00015,
|
990 |
+
"epoch": 2.36
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"current_steps": 165,
|
994 |
+
"loss": 0.7904,
|
995 |
+
"learning_rate": 0.00014727272727272725,
|
996 |
+
"epoch": 2.37
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"current_steps": 166,
|
1000 |
+
"loss": 0.7036,
|
1001 |
+
"learning_rate": 0.00014454545454545452,
|
1002 |
+
"epoch": 2.39
|
1003 |
+
},
|
1004 |
+
{
|
1005 |
+
"current_steps": 167,
|
1006 |
+
"loss": 0.5043,
|
1007 |
+
"learning_rate": 0.0001418181818181818,
|
1008 |
+
"epoch": 2.4
|
1009 |
+
},
|
1010 |
+
{
|
1011 |
+
"current_steps": 168,
|
1012 |
+
"loss": 0.7245,
|
1013 |
+
"learning_rate": 0.00013909090909090908,
|
1014 |
+
"epoch": 2.41
|
1015 |
+
},
|
1016 |
+
{
|
1017 |
+
"current_steps": 169,
|
1018 |
+
"loss": 0.7101,
|
1019 |
+
"learning_rate": 0.00013636363636363634,
|
1020 |
+
"epoch": 2.43
|
1021 |
+
},
|
1022 |
+
{
|
1023 |
+
"current_steps": 170,
|
1024 |
+
"loss": 0.3898,
|
1025 |
+
"learning_rate": 0.0001336363636363636,
|
1026 |
+
"epoch": 2.44
|
1027 |
+
},
|
1028 |
+
{
|
1029 |
+
"current_steps": 171,
|
1030 |
+
"loss": 0.4639,
|
1031 |
+
"learning_rate": 0.0001309090909090909,
|
1032 |
+
"epoch": 2.46
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"current_steps": 172,
|
1036 |
+
"loss": 0.686,
|
1037 |
+
"learning_rate": 0.00012818181818181817,
|
1038 |
+
"epoch": 2.47
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"current_steps": 173,
|
1042 |
+
"loss": 0.4952,
|
1043 |
+
"learning_rate": 0.00012545454545454543,
|
1044 |
+
"epoch": 2.49
|
1045 |
+
},
|
1046 |
+
{
|
1047 |
+
"current_steps": 174,
|
1048 |
+
"loss": 0.2727,
|
1049 |
+
"learning_rate": 0.00012272727272727272,
|
1050 |
+
"epoch": 2.5
|
1051 |
+
},
|
1052 |
+
{
|
1053 |
+
"current_steps": 175,
|
1054 |
+
"loss": 0.3428,
|
1055 |
+
"learning_rate": 0.00011999999999999999,
|
1056 |
+
"epoch": 2.51
|
1057 |
+
},
|
1058 |
+
{
|
1059 |
+
"current_steps": 176,
|
1060 |
+
"loss": 0.253,
|
1061 |
+
"learning_rate": 0.00011727272727272727,
|
1062 |
+
"epoch": 2.53
|
1063 |
+
},
|
1064 |
+
{
|
1065 |
+
"current_steps": 177,
|
1066 |
+
"loss": 0.5778,
|
1067 |
+
"learning_rate": 0.00011454545454545453,
|
1068 |
+
"epoch": 2.54
|
1069 |
+
},
|
1070 |
+
{
|
1071 |
+
"current_steps": 178,
|
1072 |
+
"loss": 0.639,
|
1073 |
+
"learning_rate": 0.00011181818181818181,
|
1074 |
+
"epoch": 2.56
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"current_steps": 179,
|
1078 |
+
"loss": 0.7327,
|
1079 |
+
"learning_rate": 0.00010909090909090908,
|
1080 |
+
"epoch": 2.57
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"current_steps": 180,
|
1084 |
+
"loss": 0.4956,
|
1085 |
+
"learning_rate": 0.00010636363636363636,
|
1086 |
+
"epoch": 2.59
|
1087 |
+
},
|
1088 |
+
{
|
1089 |
+
"current_steps": 181,
|
1090 |
+
"loss": 0.364,
|
1091 |
+
"learning_rate": 0.00010363636363636362,
|
1092 |
+
"epoch": 2.6
|
1093 |
+
},
|
1094 |
+
{
|
1095 |
+
"current_steps": 182,
|
1096 |
+
"loss": 0.4497,
|
1097 |
+
"learning_rate": 0.0001009090909090909,
|
1098 |
+
"epoch": 2.61
|
1099 |
+
},
|
1100 |
+
{
|
1101 |
+
"current_steps": 183,
|
1102 |
+
"loss": 0.7653,
|
1103 |
+
"learning_rate": 9.818181818181817e-05,
|
1104 |
+
"epoch": 2.63
|
1105 |
+
},
|
1106 |
+
{
|
1107 |
+
"current_steps": 184,
|
1108 |
+
"loss": 0.6586,
|
1109 |
+
"learning_rate": 9.545454545454545e-05,
|
1110 |
+
"epoch": 2.64
|
1111 |
+
},
|
1112 |
+
{
|
1113 |
+
"current_steps": 185,
|
1114 |
+
"loss": 0.4404,
|
1115 |
+
"learning_rate": 9.272727272727271e-05,
|
1116 |
+
"epoch": 2.66
|
1117 |
+
},
|
1118 |
+
{
|
1119 |
+
"current_steps": 186,
|
1120 |
+
"loss": 0.7484,
|
1121 |
+
"learning_rate": 8.999999999999999e-05,
|
1122 |
+
"epoch": 2.67
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"current_steps": 187,
|
1126 |
+
"loss": 0.6176,
|
1127 |
+
"learning_rate": 8.727272727272726e-05,
|
1128 |
+
"epoch": 2.69
|
1129 |
+
},
|
1130 |
+
{
|
1131 |
+
"current_steps": 188,
|
1132 |
+
"loss": 0.7404,
|
1133 |
+
"learning_rate": 8.454545454545454e-05,
|
1134 |
+
"epoch": 2.7
|
1135 |
+
},
|
1136 |
+
{
|
1137 |
+
"current_steps": 189,
|
1138 |
+
"loss": 0.8124,
|
1139 |
+
"learning_rate": 8.18181818181818e-05,
|
1140 |
+
"epoch": 2.71
|
1141 |
+
},
|
1142 |
+
{
|
1143 |
+
"current_steps": 190,
|
1144 |
+
"loss": 0.6525,
|
1145 |
+
"learning_rate": 7.909090909090908e-05,
|
1146 |
+
"epoch": 2.73
|
1147 |
+
},
|
1148 |
+
{
|
1149 |
+
"current_steps": 191,
|
1150 |
+
"loss": 0.5986,
|
1151 |
+
"learning_rate": 7.636363636363635e-05,
|
1152 |
+
"epoch": 2.74
|
1153 |
+
},
|
1154 |
+
{
|
1155 |
+
"current_steps": 192,
|
1156 |
+
"loss": 0.294,
|
1157 |
+
"learning_rate": 7.363636363636363e-05,
|
1158 |
+
"epoch": 2.76
|
1159 |
+
},
|
1160 |
+
{
|
1161 |
+
"current_steps": 193,
|
1162 |
+
"loss": 0.2973,
|
1163 |
+
"learning_rate": 7.09090909090909e-05,
|
1164 |
+
"epoch": 2.77
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"current_steps": 194,
|
1168 |
+
"loss": 0.3589,
|
1169 |
+
"learning_rate": 6.818181818181817e-05,
|
1170 |
+
"epoch": 2.79
|
1171 |
+
},
|
1172 |
+
{
|
1173 |
+
"current_steps": 195,
|
1174 |
+
"loss": 0.78,
|
1175 |
+
"learning_rate": 6.545454545454545e-05,
|
1176 |
+
"epoch": 2.8
|
1177 |
+
},
|
1178 |
+
{
|
1179 |
+
"current_steps": 196,
|
1180 |
+
"loss": 0.4056,
|
1181 |
+
"learning_rate": 6.272727272727272e-05,
|
1182 |
+
"epoch": 2.81
|
1183 |
+
},
|
1184 |
+
{
|
1185 |
+
"current_steps": 197,
|
1186 |
+
"loss": 0.5843,
|
1187 |
+
"learning_rate": 5.9999999999999995e-05,
|
1188 |
+
"epoch": 2.83
|
1189 |
+
},
|
1190 |
+
{
|
1191 |
+
"current_steps": 198,
|
1192 |
+
"loss": 0.5745,
|
1193 |
+
"learning_rate": 5.727272727272727e-05,
|
1194 |
+
"epoch": 2.84
|
1195 |
+
},
|
1196 |
+
{
|
1197 |
+
"current_steps": 199,
|
1198 |
+
"loss": 0.2763,
|
1199 |
+
"learning_rate": 5.454545454545454e-05,
|
1200 |
+
"epoch": 2.86
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"current_steps": 200,
|
1204 |
+
"loss": 0.5494,
|
1205 |
+
"learning_rate": 5.181818181818181e-05,
|
1206 |
+
"epoch": 2.87
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"current_steps": 201,
|
1210 |
+
"loss": 0.4962,
|
1211 |
+
"learning_rate": 4.9090909090909084e-05,
|
1212 |
+
"epoch": 2.89
|
1213 |
+
},
|
1214 |
+
{
|
1215 |
+
"current_steps": 202,
|
1216 |
+
"loss": 0.5864,
|
1217 |
+
"learning_rate": 4.6363636363636356e-05,
|
1218 |
+
"epoch": 2.9
|
1219 |
+
},
|
1220 |
+
{
|
1221 |
+
"current_steps": 203,
|
1222 |
+
"loss": 0.3165,
|
1223 |
+
"learning_rate": 4.363636363636363e-05,
|
1224 |
+
"epoch": 2.91
|
1225 |
+
},
|
1226 |
+
{
|
1227 |
+
"current_steps": 204,
|
1228 |
+
"loss": 0.5969,
|
1229 |
+
"learning_rate": 4.09090909090909e-05,
|
1230 |
+
"epoch": 2.93
|
1231 |
+
},
|
1232 |
+
{
|
1233 |
+
"current_steps": 205,
|
1234 |
+
"loss": 0.3003,
|
1235 |
+
"learning_rate": 3.8181818181818174e-05,
|
1236 |
+
"epoch": 2.94
|
1237 |
+
},
|
1238 |
+
{
|
1239 |
+
"current_steps": 206,
|
1240 |
+
"loss": 0.6398,
|
1241 |
+
"learning_rate": 3.545454545454545e-05,
|
1242 |
+
"epoch": 2.96
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"current_steps": 207,
|
1246 |
+
"loss": 0.604,
|
1247 |
+
"learning_rate": 3.2727272727272725e-05,
|
1248 |
+
"epoch": 2.97
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"current_steps": 208,
|
1252 |
+
"loss": 0.6389,
|
1253 |
+
"learning_rate": 2.9999999999999997e-05,
|
1254 |
+
"epoch": 2.99
|
1255 |
+
},
|
1256 |
+
{
|
1257 |
+
"current_steps": 209,
|
1258 |
+
"loss": 0.5865,
|
1259 |
+
"learning_rate": 2.727272727272727e-05,
|
1260 |
+
"epoch": 3.0
|
1261 |
+
},
|
1262 |
+
{
|
1263 |
+
"current_steps": 209,
|
1264 |
+
"loss": 0.5865,
|
1265 |
+
"learning_rate": 2.727272727272727e-05,
|
1266 |
+
"epoch": 3.0
|
1267 |
+
}
|
1268 |
+
]
|
training_graph.png
ADDED
training_log.json
ADDED
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"base_model_name": "HuggingFaceH4_zephyr-7b-alpha",
|
3 |
+
"base_model_class": "MistralForCausalLM",
|
4 |
+
"base_loaded_in_4bit": true,
|
5 |
+
"base_loaded_in_8bit": false,
|
6 |
+
"projections": "q, v",
|
7 |
+
"loss": 0.5865,
|
8 |
+
"learning_rate": 2.727272727272727e-05,
|
9 |
+
"epoch": 3.0,
|
10 |
+
"current_steps": 209,
|
11 |
+
"current_steps_adjusted": 209,
|
12 |
+
"epoch_adjusted": 3.0,
|
13 |
+
"train_runtime": 152.7978,
|
14 |
+
"train_samples_per_second": 5.478,
|
15 |
+
"train_steps_per_second": 1.374,
|
16 |
+
"total_flos": 9159217237721088.0,
|
17 |
+
"train_loss": 1.2863307027589708
|
18 |
+
}
|
training_parameters.json
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"lora_name": "zephyr7b_prompts",
|
3 |
+
"always_override": false,
|
4 |
+
"save_steps": 0.0,
|
5 |
+
"micro_batch_size": 4,
|
6 |
+
"batch_size": 0,
|
7 |
+
"epochs": 3.0,
|
8 |
+
"learning_rate": "3e-4",
|
9 |
+
"lr_scheduler_type": "linear",
|
10 |
+
"lora_rank": 32,
|
11 |
+
"lora_alpha": 64,
|
12 |
+
"lora_dropout": 0.05,
|
13 |
+
"cutoff_len": 256,
|
14 |
+
"dataset": "None",
|
15 |
+
"eval_dataset": "None",
|
16 |
+
"format": "None",
|
17 |
+
"eval_steps": 100.0,
|
18 |
+
"raw_text_file": "singles",
|
19 |
+
"higher_rank_limit": false,
|
20 |
+
"warmup_steps": 100.0,
|
21 |
+
"optimizer": "adamw_torch",
|
22 |
+
"hard_cut_string": "end;\\n",
|
23 |
+
"train_only_after": "",
|
24 |
+
"stop_at_loss": 0,
|
25 |
+
"add_eos_token": false,
|
26 |
+
"min_chars": 0.0,
|
27 |
+
"report_to": "None",
|
28 |
+
"precize_slicing_overlap": true,
|
29 |
+
"add_eos_token_type": "Every Block",
|
30 |
+
"save_steps_under_loss": 1.8,
|
31 |
+
"add_bos_token": true,
|
32 |
+
"training_projection": "q-v",
|
33 |
+
"sliding_window": false,
|
34 |
+
"warmup_ratio": 0,
|
35 |
+
"grad_accumulation": 1,
|
36 |
+
"neft_noise_alpha": 0
|
37 |
+
}
|
training_prompt.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"template_type": "raw_text"
|
3 |
+
}
|