Training in progress, step 100, checkpoint
Browse files
checkpoint-100/adapter_config.json
CHANGED
@@ -20,10 +20,10 @@
|
|
20 |
"rank_pattern": {},
|
21 |
"revision": null,
|
22 |
"target_modules": [
|
23 |
-
"o_proj",
|
24 |
-
"k_proj",
|
25 |
"q_proj",
|
26 |
-
"
|
|
|
|
|
27 |
],
|
28 |
"task_type": "CAUSAL_LM",
|
29 |
"use_dora": false,
|
|
|
20 |
"rank_pattern": {},
|
21 |
"revision": null,
|
22 |
"target_modules": [
|
|
|
|
|
23 |
"q_proj",
|
24 |
+
"k_proj",
|
25 |
+
"v_proj",
|
26 |
+
"o_proj"
|
27 |
],
|
28 |
"task_type": "CAUSAL_LM",
|
29 |
"use_dora": false,
|
checkpoint-100/adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 54560368
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9022a41c6a7725ac3af5b0936fc7e06fc2895c4a8da259582c34fce1136a0fd4
|
3 |
size 54560368
|
checkpoint-100/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 109267450
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d3821e0c5652183b68ba5df1fec03c5f08c944e6453f131c4b4c0eb3216f0912
|
3 |
size 109267450
|
checkpoint-100/trainer_state.json
CHANGED
@@ -1,6 +1,6 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "./zephyr/
|
4 |
"epoch": 0.684931506849315,
|
5 |
"eval_steps": 50,
|
6 |
"global_step": 100,
|
@@ -10,42 +10,42 @@
|
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.14,
|
13 |
-
"grad_norm":
|
14 |
-
"kl":
|
15 |
"learning_rate": 0.00018,
|
16 |
-
"logps/chosen": -
|
17 |
-
"logps/rejected": -
|
18 |
-
"loss": 0.
|
19 |
-
"rewards/chosen":
|
20 |
-
"rewards/margins":
|
21 |
-
"rewards/rejected":
|
22 |
"step": 20
|
23 |
},
|
24 |
{
|
25 |
"epoch": 0.27,
|
26 |
"grad_norm": 0.0,
|
27 |
-
"kl":
|
28 |
"learning_rate": 0.00015142857142857143,
|
29 |
-
"logps/chosen": -
|
30 |
-
"logps/rejected": -
|
31 |
-
"loss": 0.
|
32 |
-
"rewards/chosen": -
|
33 |
-
"rewards/margins":
|
34 |
-
"rewards/rejected": -
|
35 |
"step": 40
|
36 |
},
|
37 |
{
|
38 |
"epoch": 0.34,
|
39 |
"eval_kl": 0.0,
|
40 |
-
"eval_logps/chosen": -
|
41 |
-
"eval_logps/rejected": -
|
42 |
-
"eval_loss": 0.
|
43 |
-
"eval_rewards/chosen": -
|
44 |
-
"eval_rewards/margins":
|
45 |
-
"eval_rewards/rejected": -
|
46 |
-
"eval_runtime": 138.
|
47 |
-
"eval_samples_per_second": 2.
|
48 |
-
"eval_steps_per_second": 0.
|
49 |
"step": 50
|
50 |
},
|
51 |
{
|
@@ -53,12 +53,12 @@
|
|
53 |
"grad_norm": 0.0,
|
54 |
"kl": 0.0,
|
55 |
"learning_rate": 0.00012285714285714287,
|
56 |
-
"logps/chosen": -
|
57 |
-
"logps/rejected": -
|
58 |
-
"loss": 0.
|
59 |
-
"rewards/chosen": -
|
60 |
-
"rewards/margins":
|
61 |
-
"rewards/rejected": -
|
62 |
"step": 60
|
63 |
},
|
64 |
{
|
@@ -66,12 +66,12 @@
|
|
66 |
"grad_norm": 0.0,
|
67 |
"kl": 0.0,
|
68 |
"learning_rate": 9.428571428571429e-05,
|
69 |
-
"logps/chosen": -
|
70 |
-
"logps/rejected": -
|
71 |
-
"loss": 0.
|
72 |
-
"rewards/chosen": -
|
73 |
-
"rewards/margins":
|
74 |
-
"rewards/rejected": -
|
75 |
"step": 80
|
76 |
},
|
77 |
{
|
@@ -79,26 +79,26 @@
|
|
79 |
"grad_norm": 0.0,
|
80 |
"kl": 0.0,
|
81 |
"learning_rate": 6.571428571428571e-05,
|
82 |
-
"logps/chosen": -
|
83 |
-
"logps/rejected": -
|
84 |
-
"loss": 0.
|
85 |
-
"rewards/chosen": -
|
86 |
-
"rewards/margins": -
|
87 |
-
"rewards/rejected": -
|
88 |
"step": 100
|
89 |
},
|
90 |
{
|
91 |
"epoch": 0.68,
|
92 |
"eval_kl": 0.0,
|
93 |
-
"eval_logps/chosen": -
|
94 |
-
"eval_logps/rejected": -
|
95 |
-
"eval_loss": 0.
|
96 |
-
"eval_rewards/chosen": -
|
97 |
-
"eval_rewards/margins":
|
98 |
-
"eval_rewards/rejected": -
|
99 |
-
"eval_runtime": 138.
|
100 |
-
"eval_samples_per_second": 2.
|
101 |
-
"eval_steps_per_second": 0.
|
102 |
"step": 100
|
103 |
}
|
104 |
],
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.544333279132843,
|
3 |
+
"best_model_checkpoint": "./zephyr/09-04-24-Weni-WeniGPT-Agents-Zephyr-1.0.9-KTO_Hyperparameter search, altering desired and undesired weights for KTO task.-2_max_steps-145_batch_16_2024-04-09_ppid_9/checkpoint-100",
|
4 |
"epoch": 0.684931506849315,
|
5 |
"eval_steps": 50,
|
6 |
"global_step": 100,
|
|
|
10 |
"log_history": [
|
11 |
{
|
12 |
"epoch": 0.14,
|
13 |
+
"grad_norm": 0.0,
|
14 |
+
"kl": 0.9986292123794556,
|
15 |
"learning_rate": 0.00018,
|
16 |
+
"logps/chosen": -876.9887084960938,
|
17 |
+
"logps/rejected": -891.7728271484375,
|
18 |
+
"loss": 0.5066,
|
19 |
+
"rewards/chosen": -59.218875885009766,
|
20 |
+
"rewards/margins": 1.2524851560592651,
|
21 |
+
"rewards/rejected": -60.5962028503418,
|
22 |
"step": 20
|
23 |
},
|
24 |
{
|
25 |
"epoch": 0.27,
|
26 |
"grad_norm": 0.0,
|
27 |
+
"kl": 0.0,
|
28 |
"learning_rate": 0.00015142857142857143,
|
29 |
+
"logps/chosen": -1901.8116455078125,
|
30 |
+
"logps/rejected": -1940.32470703125,
|
31 |
+
"loss": 0.5606,
|
32 |
+
"rewards/chosen": -161.08740234375,
|
33 |
+
"rewards/margins": 2.580688953399658,
|
34 |
+
"rewards/rejected": -164.8165283203125,
|
35 |
"step": 40
|
36 |
},
|
37 |
{
|
38 |
"epoch": 0.34,
|
39 |
"eval_kl": 0.0,
|
40 |
+
"eval_logps/chosen": -1797.90625,
|
41 |
+
"eval_logps/rejected": -1778.623291015625,
|
42 |
+
"eval_loss": 0.5443333983421326,
|
43 |
+
"eval_rewards/chosen": -151.05047607421875,
|
44 |
+
"eval_rewards/margins": 2.7896132469177246,
|
45 |
+
"eval_rewards/rejected": -151.03567504882812,
|
46 |
+
"eval_runtime": 138.2181,
|
47 |
+
"eval_samples_per_second": 2.17,
|
48 |
+
"eval_steps_per_second": 0.543,
|
49 |
"step": 50
|
50 |
},
|
51 |
{
|
|
|
53 |
"grad_norm": 0.0,
|
54 |
"kl": 0.0,
|
55 |
"learning_rate": 0.00012285714285714287,
|
56 |
+
"logps/chosen": -1852.9610595703125,
|
57 |
+
"logps/rejected": -1893.4521484375,
|
58 |
+
"loss": 0.5211,
|
59 |
+
"rewards/chosen": -157.99769592285156,
|
60 |
+
"rewards/margins": 3.8418266773223877,
|
61 |
+
"rewards/rejected": -160.36117553710938,
|
62 |
"step": 60
|
63 |
},
|
64 |
{
|
|
|
66 |
"grad_norm": 0.0,
|
67 |
"kl": 0.0,
|
68 |
"learning_rate": 9.428571428571429e-05,
|
69 |
+
"logps/chosen": -2051.344482421875,
|
70 |
+
"logps/rejected": -1892.0853271484375,
|
71 |
+
"loss": 0.5714,
|
72 |
+
"rewards/chosen": -176.026611328125,
|
73 |
+
"rewards/margins": -20.468326568603516,
|
74 |
+
"rewards/rejected": -161.7165069580078,
|
75 |
"step": 80
|
76 |
},
|
77 |
{
|
|
|
79 |
"grad_norm": 0.0,
|
80 |
"kl": 0.0,
|
81 |
"learning_rate": 6.571428571428571e-05,
|
82 |
+
"logps/chosen": -2035.6630859375,
|
83 |
+
"logps/rejected": -1939.001220703125,
|
84 |
+
"loss": 0.4995,
|
85 |
+
"rewards/chosen": -175.16189575195312,
|
86 |
+
"rewards/margins": -12.196085929870605,
|
87 |
+
"rewards/rejected": -165.23004150390625,
|
88 |
"step": 100
|
89 |
},
|
90 |
{
|
91 |
"epoch": 0.68,
|
92 |
"eval_kl": 0.0,
|
93 |
+
"eval_logps/chosen": -1859.810791015625,
|
94 |
+
"eval_logps/rejected": -1831.84814453125,
|
95 |
+
"eval_loss": 0.544333279132843,
|
96 |
+
"eval_rewards/chosen": -157.24095153808594,
|
97 |
+
"eval_rewards/margins": 2.0474841594696045,
|
98 |
+
"eval_rewards/rejected": -156.35816955566406,
|
99 |
+
"eval_runtime": 138.0332,
|
100 |
+
"eval_samples_per_second": 2.173,
|
101 |
+
"eval_steps_per_second": 0.543,
|
102 |
"step": 100
|
103 |
}
|
104 |
],
|
checkpoint-100/training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5688
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:610ac0203a49e0c4734b308d39dc63c437e14685b90417052adfcc16864e15eb
|
3 |
size 5688
|