condition-3-zh-5k-seed456: trained adapter and metrics

#24
by rafaym - opened
tiny-aya-base/condition-3-zh-5k-seed456/adapter_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alora_invocation_tokens": null,
3
+ "alpha_pattern": {},
4
+ "arrow_config": null,
5
+ "auto_mapping": {
6
+ "base_model_class": "Cohere2ForCausalLM",
7
+ "parent_library": "unsloth_compiled_module_cohere2",
8
+ "unsloth_fixed": true
9
+ },
10
+ "base_model_name_or_path": "CohereLabs/tiny-aya-base",
11
+ "bias": "none",
12
+ "corda_config": null,
13
+ "ensure_weight_tying": false,
14
+ "eva_config": null,
15
+ "exclude_modules": null,
16
+ "fan_in_fan_out": false,
17
+ "inference_mode": true,
18
+ "init_lora_weights": true,
19
+ "kasa_config": null,
20
+ "layer_replication": null,
21
+ "layers_pattern": null,
22
+ "layers_to_transform": null,
23
+ "loftq_config": {},
24
+ "lora_alpha": 32,
25
+ "lora_bias": false,
26
+ "lora_dropout": 0.0,
27
+ "lora_ga_config": null,
28
+ "megatron_config": null,
29
+ "megatron_core": "megatron.core",
30
+ "modules_to_save": null,
31
+ "monteclora_config": null,
32
+ "peft_type": "LORA",
33
+ "peft_version": "0.21.2",
34
+ "qalora_group_size": 16,
35
+ "r": 16,
36
+ "rank_pattern": {},
37
+ "revision": null,
38
+ "target_modules": [
39
+ "down_proj",
40
+ "k_proj",
41
+ "gate_proj",
42
+ "q_proj",
43
+ "o_proj",
44
+ "v_proj",
45
+ "up_proj"
46
+ ],
47
+ "target_parameters": null,
48
+ "task_type": "CAUSAL_LM",
49
+ "trainable_token_indices": null,
50
+ "use_bdlora": null,
51
+ "use_dora": false,
52
+ "use_qalora": false,
53
+ "use_rslora": false,
54
+ "velora_config": null
55
+ }
tiny-aya-base/condition-3-zh-5k-seed456/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ed1da542d660f94f8778b48f91e3dea98975aee68b6e059bd4d64e31b7e7eeaa
3
+ size 120981200
tiny-aya-base/condition-3-zh-5k-seed456/training_metrics.json ADDED
@@ -0,0 +1,238 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_config_name": "tiny-aya-base",
3
+ "condition_name": "condition-3-zh-5k",
4
+ "seed": 456,
5
+ "output_name": "condition-3-zh-5k-seed456",
6
+ "train_result": {
7
+ "train_runtime": 2892.6829,
8
+ "train_samples_per_second": 1.556,
9
+ "train_steps_per_second": 0.097,
10
+ "total_flos": 7.8418920700117e+16,
11
+ "train_loss": 1.1631332153969622,
12
+ "epoch": 1.0
13
+ },
14
+ "environment": {
15
+ "python": "3.13.16",
16
+ "cuda": "13.0",
17
+ "gpu": "Tesla T4",
18
+ "packages": {
19
+ "unsloth": "2026.11.3",
20
+ "unsloth_zoo": "2026.11.4",
21
+ "transformers": "4.56.2",
22
+ "trl": "0.22.2",
23
+ "peft": "0.21.2",
24
+ "torch": "2.11.0+cu130",
25
+ "accelerate": "1.15.0",
26
+ "bitsandbytes": "0.50.2",
27
+ "datasets": "4.8.5",
28
+ "huggingface_hub": "0.36.2"
29
+ }
30
+ },
31
+ "log_history": [
32
+ {
33
+ "loss": 1.2787,
34
+ "grad_norm": 0.12460597604513168,
35
+ "learning_rate": 0.00012,
36
+ "epoch": 0.03546099290780142,
37
+ "step": 10
38
+ },
39
+ {
40
+ "loss": 1.3017,
41
+ "grad_norm": 0.15038713812828064,
42
+ "learning_rate": 0.00019988926445681492,
43
+ "epoch": 0.07092198581560284,
44
+ "step": 20
45
+ },
46
+ {
47
+ "loss": 1.261,
48
+ "grad_norm": 0.1366133987903595,
49
+ "learning_rate": 0.0001986463043917528,
50
+ "epoch": 0.10638297872340426,
51
+ "step": 30
52
+ },
53
+ {
54
+ "loss": 1.181,
55
+ "grad_norm": 0.16550225019454956,
56
+ "learning_rate": 0.00019603921063437793,
57
+ "epoch": 0.14184397163120568,
58
+ "step": 40
59
+ },
60
+ {
61
+ "loss": 1.1851,
62
+ "grad_norm": 0.14056502282619476,
63
+ "learning_rate": 0.0001921040354671897,
64
+ "epoch": 0.1773049645390071,
65
+ "step": 50
66
+ },
67
+ {
68
+ "loss": 1.2414,
69
+ "grad_norm": 0.18625281751155853,
70
+ "learning_rate": 0.00018689519659051467,
71
+ "epoch": 0.2127659574468085,
72
+ "step": 60
73
+ },
74
+ {
75
+ "loss": 1.1097,
76
+ "grad_norm": 0.16673700511455536,
77
+ "learning_rate": 0.00018048472460553257,
78
+ "epoch": 0.24822695035460993,
79
+ "step": 70
80
+ },
81
+ {
82
+ "loss": 1.1885,
83
+ "grad_norm": 0.1709842085838318,
84
+ "learning_rate": 0.00017296126693671884,
85
+ "epoch": 0.28368794326241137,
86
+ "step": 80
87
+ },
88
+ {
89
+ "loss": 1.1792,
90
+ "grad_norm": 0.14952898025512695,
91
+ "learning_rate": 0.00016442886196799464,
92
+ "epoch": 0.3191489361702128,
93
+ "step": 90
94
+ },
95
+ {
96
+ "loss": 1.1916,
97
+ "grad_norm": 0.12436094880104065,
98
+ "learning_rate": 0.00015500550034448413,
99
+ "epoch": 0.3546099290780142,
100
+ "step": 100
101
+ },
102
+ {
103
+ "loss": 1.0631,
104
+ "grad_norm": 0.14213046431541443,
105
+ "learning_rate": 0.00014482149333496454,
106
+ "epoch": 0.3900709219858156,
107
+ "step": 110
108
+ },
109
+ {
110
+ "loss": 1.0074,
111
+ "grad_norm": 0.14886945486068726,
112
+ "learning_rate": 0.0001340176708181637,
113
+ "epoch": 0.425531914893617,
114
+ "step": 120
115
+ },
116
+ {
117
+ "loss": 1.1875,
118
+ "grad_norm": 0.1696494072675705,
119
+ "learning_rate": 0.00012274343381211066,
120
+ "epoch": 0.46099290780141844,
121
+ "step": 130
122
+ },
123
+ {
124
+ "loss": 1.1347,
125
+ "grad_norm": 0.15974335372447968,
126
+ "learning_rate": 0.00011115468847720245,
127
+ "epoch": 0.49645390070921985,
128
+ "step": 140
129
+ },
130
+ {
131
+ "loss": 1.1425,
132
+ "grad_norm": 0.1326570063829422,
133
+ "learning_rate": 9.941169016269379e-05,
134
+ "epoch": 0.5319148936170213,
135
+ "step": 150
136
+ },
137
+ {
138
+ "loss": 1.1588,
139
+ "grad_norm": 0.14309875667095184,
140
+ "learning_rate": 8.767682731028415e-05,
141
+ "epoch": 0.5673758865248227,
142
+ "step": 160
143
+ },
144
+ {
145
+ "loss": 1.1298,
146
+ "grad_norm": 0.17677722871303558,
147
+ "learning_rate": 7.611237586016557e-05,
148
+ "epoch": 0.6028368794326241,
149
+ "step": 170
150
+ },
151
+ {
152
+ "loss": 1.1085,
153
+ "grad_norm": 0.13328783214092255,
154
+ "learning_rate": 6.487825521280109e-05,
155
+ "epoch": 0.6382978723404256,
156
+ "step": 180
157
+ },
158
+ {
159
+ "loss": 1.1626,
160
+ "grad_norm": 0.16187146306037903,
161
+ "learning_rate": 5.4129816778190936e-05,
162
+ "epoch": 0.6737588652482269,
163
+ "step": 190
164
+ },
165
+ {
166
+ "loss": 1.2559,
167
+ "grad_norm": 0.1936364471912384,
168
+ "learning_rate": 4.401569569374668e-05,
169
+ "epoch": 0.7092198581560284,
170
+ "step": 200
171
+ },
172
+ {
173
+ "loss": 1.1318,
174
+ "grad_norm": 0.134074866771698,
175
+ "learning_rate": 3.467575541836305e-05,
176
+ "epoch": 0.7446808510638298,
177
+ "step": 210
178
+ },
179
+ {
180
+ "loss": 1.1784,
181
+ "grad_norm": 0.1537429839372635,
182
+ "learning_rate": 2.6239153625937784e-05,
183
+ "epoch": 0.7801418439716312,
184
+ "step": 220
185
+ },
186
+ {
187
+ "loss": 1.0611,
188
+ "grad_norm": 0.13412530720233917,
189
+ "learning_rate": 1.882255614419376e-05,
190
+ "epoch": 0.8156028368794326,
191
+ "step": 230
192
+ },
193
+ {
194
+ "loss": 1.0693,
195
+ "grad_norm": 0.1476028859615326,
196
+ "learning_rate": 1.2528523637410838e-05,
197
+ "epoch": 0.851063829787234,
198
+ "step": 240
199
+ },
200
+ {
201
+ "loss": 1.1029,
202
+ "grad_norm": 0.13800477981567383,
203
+ "learning_rate": 7.4440933428779e-06,
204
+ "epoch": 0.8865248226950354,
205
+ "step": 250
206
+ },
207
+ {
208
+ "loss": 1.1354,
209
+ "grad_norm": 0.13396550714969635,
210
+ "learning_rate": 3.6395754735699894e-06,
211
+ "epoch": 0.9219858156028369,
212
+ "step": 260
213
+ },
214
+ {
215
+ "loss": 1.2356,
216
+ "grad_norm": 0.16255514323711395,
217
+ "learning_rate": 1.1675809310361497e-06,
218
+ "epoch": 0.9574468085106383,
219
+ "step": 270
220
+ },
221
+ {
222
+ "loss": 1.1756,
223
+ "grad_norm": 0.20377686619758606,
224
+ "learning_rate": 6.229377380218005e-08,
225
+ "epoch": 0.9929078014184397,
226
+ "step": 280
227
+ },
228
+ {
229
+ "train_runtime": 2892.6829,
230
+ "train_samples_per_second": 1.556,
231
+ "train_steps_per_second": 0.097,
232
+ "total_flos": 7.8418920700117e+16,
233
+ "train_loss": 1.1631332153969622,
234
+ "epoch": 1.0,
235
+ "step": 282
236
+ }
237
+ ]
238
+ }