Training in progress, step 50
Browse files- adapter_config.json +3 -3
- adapter_model.safetensors +1 -1
- metrics.json +1 -18
- state.json +17 -272
- tokenizer_config.json +4 -0
- training_args.bin +1 -1
adapter_config.json
CHANGED
@@ -20,13 +20,13 @@
|
|
20 |
"rank_pattern": {},
|
21 |
"revision": null,
|
22 |
"target_modules": [
|
23 |
-
"k_proj",
|
24 |
"o_proj",
|
25 |
"q_proj",
|
26 |
-
"up_proj",
|
27 |
"down_proj",
|
28 |
"gate_proj",
|
29 |
-
"v_proj"
|
|
|
|
|
30 |
],
|
31 |
"task_type": "CAUSAL_LM",
|
32 |
"use_dora": false,
|
|
|
20 |
"rank_pattern": {},
|
21 |
"revision": null,
|
22 |
"target_modules": [
|
|
|
23 |
"o_proj",
|
24 |
"q_proj",
|
|
|
25 |
"down_proj",
|
26 |
"gate_proj",
|
27 |
+
"v_proj",
|
28 |
+
"k_proj",
|
29 |
+
"up_proj"
|
30 |
],
|
31 |
"task_type": "CAUSAL_LM",
|
32 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 35668592
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2bf964742964d2ec503f25da4ead7a6c8a5691fc8d1cb60aa46b3a6fe8d444f8
|
3 |
size 35668592
|
metrics.json
CHANGED
@@ -1,18 +1 @@
|
|
1 |
-
{"Step":
|
2 |
-
{"Step":40,"eval_loss":1.8370074034,"eval_runtime":1516.9668,"eval_samples_per_second":3.296,"eval_steps_per_second":0.412,"epoch":0.064}
|
3 |
-
{"Step":60,"eval_loss":1.6439222097,"eval_runtime":1500.7581,"eval_samples_per_second":3.332,"eval_steps_per_second":0.416,"epoch":0.096}
|
4 |
-
{"Step":80,"eval_loss":1.550334692,"eval_runtime":1506.1101,"eval_samples_per_second":3.32,"eval_steps_per_second":0.415,"epoch":0.128}
|
5 |
-
{"Step":100,"eval_loss":1.519143939,"eval_runtime":1499.8012,"eval_samples_per_second":3.334,"eval_steps_per_second":0.417,"epoch":0.16}
|
6 |
-
{"Step":120,"eval_loss":1.5071601868,"eval_runtime":1498.469,"eval_samples_per_second":3.337,"eval_steps_per_second":0.417,"epoch":0.192}
|
7 |
-
{"Step":140,"eval_loss":1.4998296499,"eval_runtime":1486.0248,"eval_samples_per_second":3.365,"eval_steps_per_second":0.421,"epoch":0.224}
|
8 |
-
{"Step":160,"eval_loss":1.4962984324,"eval_runtime":1491.4174,"eval_samples_per_second":3.353,"eval_steps_per_second":0.419,"epoch":0.256}
|
9 |
-
{"Step":180,"eval_loss":1.4931029081,"eval_runtime":1501.5451,"eval_samples_per_second":3.33,"eval_steps_per_second":0.416,"epoch":0.288}
|
10 |
-
{"Step":200,"eval_loss":1.4910326004,"eval_runtime":1501.9275,"eval_samples_per_second":3.329,"eval_steps_per_second":0.416,"epoch":0.32}
|
11 |
-
{"Step":220,"eval_loss":1.488992691,"eval_runtime":1498.0439,"eval_samples_per_second":3.338,"eval_steps_per_second":0.417,"epoch":0.352}
|
12 |
-
{"Step":240,"eval_loss":1.4867918491,"eval_runtime":1460.1808,"eval_samples_per_second":3.424,"eval_steps_per_second":0.428,"epoch":0.384}
|
13 |
-
{"Step":260,"eval_loss":1.4859720469,"eval_runtime":1460.3853,"eval_samples_per_second":3.424,"eval_steps_per_second":0.428,"epoch":0.416}
|
14 |
-
{"Step":280,"eval_loss":1.4846266508,"eval_runtime":1460.0793,"eval_samples_per_second":3.424,"eval_steps_per_second":0.428,"epoch":0.448}
|
15 |
-
{"Step":300,"eval_loss":1.4849002361,"eval_runtime":1472.1795,"eval_samples_per_second":3.396,"eval_steps_per_second":0.425,"epoch":0.48}
|
16 |
-
{"Step":320,"eval_loss":1.4825308323,"eval_runtime":1500.8065,"eval_samples_per_second":3.332,"eval_steps_per_second":0.416,"epoch":0.512}
|
17 |
-
{"Step":340,"eval_loss":1.4830516577,"eval_runtime":1501.5622,"eval_samples_per_second":3.33,"eval_steps_per_second":0.416,"epoch":0.544}
|
18 |
-
{"Step":360,"eval_loss":1.4818130732,"eval_runtime":1501.7616,"eval_samples_per_second":3.329,"eval_steps_per_second":0.416,"epoch":0.576}
|
|
|
1 |
+
{"Step":50,"eval_loss":1.7156720161,"eval_runtime":149.1859,"eval_samples_per_second":3.352,"eval_steps_per_second":0.422,"epoch":0.08}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
state.json
CHANGED
@@ -1,289 +1,34 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch": 0.
|
5 |
-
"eval_steps":
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
-
"epoch": 0.
|
13 |
-
"grad_norm": 0.
|
14 |
-
"learning_rate":
|
15 |
-
"loss": 1.
|
16 |
-
"step":
|
17 |
},
|
18 |
{
|
19 |
-
"epoch": 0.
|
20 |
-
"eval_loss": 1.
|
21 |
-
"eval_runtime":
|
22 |
-
"eval_samples_per_second": 3.
|
23 |
-
"eval_steps_per_second": 0.
|
24 |
-
"step":
|
25 |
-
},
|
26 |
-
{
|
27 |
-
"epoch": 0.064,
|
28 |
-
"grad_norm": 0.039251815527677536,
|
29 |
-
"learning_rate": 0.00010638297872340425,
|
30 |
-
"loss": 1.9075,
|
31 |
-
"step": 40
|
32 |
-
},
|
33 |
-
{
|
34 |
-
"epoch": 0.064,
|
35 |
-
"eval_loss": 1.8370074033737183,
|
36 |
-
"eval_runtime": 1516.9668,
|
37 |
-
"eval_samples_per_second": 3.296,
|
38 |
-
"eval_steps_per_second": 0.412,
|
39 |
-
"step": 40
|
40 |
-
},
|
41 |
-
{
|
42 |
-
"epoch": 0.096,
|
43 |
-
"grad_norm": NaN,
|
44 |
-
"learning_rate": 0.00015691489361702128,
|
45 |
-
"loss": 1.7374,
|
46 |
-
"step": 60
|
47 |
-
},
|
48 |
-
{
|
49 |
-
"epoch": 0.096,
|
50 |
-
"eval_loss": 1.643922209739685,
|
51 |
-
"eval_runtime": 1500.7581,
|
52 |
-
"eval_samples_per_second": 3.332,
|
53 |
-
"eval_steps_per_second": 0.416,
|
54 |
-
"step": 60
|
55 |
-
},
|
56 |
-
{
|
57 |
-
"epoch": 0.128,
|
58 |
-
"grad_norm": 0.10252918303012848,
|
59 |
-
"learning_rate": 0.0002101063829787234,
|
60 |
-
"loss": 1.6047,
|
61 |
-
"step": 80
|
62 |
-
},
|
63 |
-
{
|
64 |
-
"epoch": 0.128,
|
65 |
-
"eval_loss": 1.5503346920013428,
|
66 |
-
"eval_runtime": 1506.1101,
|
67 |
-
"eval_samples_per_second": 3.32,
|
68 |
-
"eval_steps_per_second": 0.415,
|
69 |
-
"step": 80
|
70 |
-
},
|
71 |
-
{
|
72 |
-
"epoch": 0.16,
|
73 |
-
"grad_norm": 0.026087351143360138,
|
74 |
-
"learning_rate": 0.0002632978723404255,
|
75 |
-
"loss": 1.539,
|
76 |
-
"step": 100
|
77 |
-
},
|
78 |
-
{
|
79 |
-
"epoch": 0.16,
|
80 |
-
"eval_loss": 1.5191439390182495,
|
81 |
-
"eval_runtime": 1499.8012,
|
82 |
-
"eval_samples_per_second": 3.334,
|
83 |
-
"eval_steps_per_second": 0.417,
|
84 |
-
"step": 100
|
85 |
-
},
|
86 |
-
{
|
87 |
-
"epoch": 0.192,
|
88 |
-
"grad_norm": 0.03835824504494667,
|
89 |
-
"learning_rate": 0.00031648936170212765,
|
90 |
-
"loss": 1.5149,
|
91 |
-
"step": 120
|
92 |
-
},
|
93 |
-
{
|
94 |
-
"epoch": 0.192,
|
95 |
-
"eval_loss": 1.5071601867675781,
|
96 |
-
"eval_runtime": 1498.469,
|
97 |
-
"eval_samples_per_second": 3.337,
|
98 |
-
"eval_steps_per_second": 0.417,
|
99 |
-
"step": 120
|
100 |
-
},
|
101 |
-
{
|
102 |
-
"epoch": 0.224,
|
103 |
-
"grad_norm": 0.03541890159249306,
|
104 |
-
"learning_rate": 0.0003696808510638298,
|
105 |
-
"loss": 1.5267,
|
106 |
-
"step": 140
|
107 |
-
},
|
108 |
-
{
|
109 |
-
"epoch": 0.224,
|
110 |
-
"eval_loss": 1.499829649925232,
|
111 |
-
"eval_runtime": 1486.0248,
|
112 |
-
"eval_samples_per_second": 3.365,
|
113 |
-
"eval_steps_per_second": 0.421,
|
114 |
-
"step": 140
|
115 |
-
},
|
116 |
-
{
|
117 |
-
"epoch": 0.256,
|
118 |
-
"grad_norm": 0.045721374452114105,
|
119 |
-
"learning_rate": 0.0004228723404255319,
|
120 |
-
"loss": 1.4927,
|
121 |
-
"step": 160
|
122 |
-
},
|
123 |
-
{
|
124 |
-
"epoch": 0.256,
|
125 |
-
"eval_loss": 1.4962984323501587,
|
126 |
-
"eval_runtime": 1491.4174,
|
127 |
-
"eval_samples_per_second": 3.353,
|
128 |
-
"eval_steps_per_second": 0.419,
|
129 |
-
"step": 160
|
130 |
-
},
|
131 |
-
{
|
132 |
-
"epoch": 0.288,
|
133 |
-
"grad_norm": 0.04470464587211609,
|
134 |
-
"learning_rate": 0.0004760638297872341,
|
135 |
-
"loss": 1.4861,
|
136 |
-
"step": 180
|
137 |
-
},
|
138 |
-
{
|
139 |
-
"epoch": 0.288,
|
140 |
-
"eval_loss": 1.4931029081344604,
|
141 |
-
"eval_runtime": 1501.5451,
|
142 |
-
"eval_samples_per_second": 3.33,
|
143 |
-
"eval_steps_per_second": 0.416,
|
144 |
-
"step": 180
|
145 |
-
},
|
146 |
-
{
|
147 |
-
"epoch": 0.32,
|
148 |
-
"grad_norm": 0.04011245444417,
|
149 |
-
"learning_rate": 0.0004967397747480735,
|
150 |
-
"loss": 1.4815,
|
151 |
-
"step": 200
|
152 |
-
},
|
153 |
-
{
|
154 |
-
"epoch": 0.32,
|
155 |
-
"eval_loss": 1.491032600402832,
|
156 |
-
"eval_runtime": 1501.9275,
|
157 |
-
"eval_samples_per_second": 3.329,
|
158 |
-
"eval_steps_per_second": 0.416,
|
159 |
-
"step": 200
|
160 |
-
},
|
161 |
-
{
|
162 |
-
"epoch": 0.352,
|
163 |
-
"grad_norm": 0.03507564216852188,
|
164 |
-
"learning_rate": 0.0004908120924718435,
|
165 |
-
"loss": 1.5256,
|
166 |
-
"step": 220
|
167 |
-
},
|
168 |
-
{
|
169 |
-
"epoch": 0.352,
|
170 |
-
"eval_loss": 1.488992691040039,
|
171 |
-
"eval_runtime": 1498.0439,
|
172 |
-
"eval_samples_per_second": 3.338,
|
173 |
-
"eval_steps_per_second": 0.417,
|
174 |
-
"step": 220
|
175 |
-
},
|
176 |
-
{
|
177 |
-
"epoch": 0.384,
|
178 |
-
"grad_norm": 0.03816624730825424,
|
179 |
-
"learning_rate": 0.0004848844101956135,
|
180 |
-
"loss": 1.4932,
|
181 |
-
"step": 240
|
182 |
-
},
|
183 |
-
{
|
184 |
-
"epoch": 0.384,
|
185 |
-
"eval_loss": 1.4867918491363525,
|
186 |
-
"eval_runtime": 1460.1808,
|
187 |
-
"eval_samples_per_second": 3.424,
|
188 |
-
"eval_steps_per_second": 0.428,
|
189 |
-
"step": 240
|
190 |
-
},
|
191 |
-
{
|
192 |
-
"epoch": 0.416,
|
193 |
-
"grad_norm": 0.03537523001432419,
|
194 |
-
"learning_rate": 0.00047895672791938357,
|
195 |
-
"loss": 1.4747,
|
196 |
-
"step": 260
|
197 |
-
},
|
198 |
-
{
|
199 |
-
"epoch": 0.416,
|
200 |
-
"eval_loss": 1.4859720468521118,
|
201 |
-
"eval_runtime": 1460.3853,
|
202 |
-
"eval_samples_per_second": 3.424,
|
203 |
-
"eval_steps_per_second": 0.428,
|
204 |
-
"step": 260
|
205 |
-
},
|
206 |
-
{
|
207 |
-
"epoch": 0.448,
|
208 |
-
"grad_norm": 0.037848278880119324,
|
209 |
-
"learning_rate": 0.0004730290456431535,
|
210 |
-
"loss": 1.4635,
|
211 |
-
"step": 280
|
212 |
-
},
|
213 |
-
{
|
214 |
-
"epoch": 0.448,
|
215 |
-
"eval_loss": 1.4846266508102417,
|
216 |
-
"eval_runtime": 1460.0793,
|
217 |
-
"eval_samples_per_second": 3.424,
|
218 |
-
"eval_steps_per_second": 0.428,
|
219 |
-
"step": 280
|
220 |
-
},
|
221 |
-
{
|
222 |
-
"epoch": 0.48,
|
223 |
-
"grad_norm": 0.038794029504060745,
|
224 |
-
"learning_rate": 0.00046710136336692356,
|
225 |
-
"loss": 1.5055,
|
226 |
-
"step": 300
|
227 |
-
},
|
228 |
-
{
|
229 |
-
"epoch": 0.48,
|
230 |
-
"eval_loss": 1.4849002361297607,
|
231 |
-
"eval_runtime": 1472.1795,
|
232 |
-
"eval_samples_per_second": 3.396,
|
233 |
-
"eval_steps_per_second": 0.425,
|
234 |
-
"step": 300
|
235 |
-
},
|
236 |
-
{
|
237 |
-
"epoch": 0.512,
|
238 |
-
"grad_norm": 0.03938526660203934,
|
239 |
-
"learning_rate": 0.00046117368109069355,
|
240 |
-
"loss": 1.4996,
|
241 |
-
"step": 320
|
242 |
-
},
|
243 |
-
{
|
244 |
-
"epoch": 0.512,
|
245 |
-
"eval_loss": 1.4825308322906494,
|
246 |
-
"eval_runtime": 1500.8065,
|
247 |
-
"eval_samples_per_second": 3.332,
|
248 |
-
"eval_steps_per_second": 0.416,
|
249 |
-
"step": 320
|
250 |
-
},
|
251 |
-
{
|
252 |
-
"epoch": 0.544,
|
253 |
-
"grad_norm": 0.041331056505441666,
|
254 |
-
"learning_rate": 0.00045524599881446355,
|
255 |
-
"loss": 1.4742,
|
256 |
-
"step": 340
|
257 |
-
},
|
258 |
-
{
|
259 |
-
"epoch": 0.544,
|
260 |
-
"eval_loss": 1.4830516576766968,
|
261 |
-
"eval_runtime": 1501.5622,
|
262 |
-
"eval_samples_per_second": 3.33,
|
263 |
-
"eval_steps_per_second": 0.416,
|
264 |
-
"step": 340
|
265 |
-
},
|
266 |
-
{
|
267 |
-
"epoch": 0.576,
|
268 |
-
"grad_norm": 0.03991026058793068,
|
269 |
-
"learning_rate": 0.0004493183165382336,
|
270 |
-
"loss": 1.4809,
|
271 |
-
"step": 360
|
272 |
-
},
|
273 |
-
{
|
274 |
-
"epoch": 0.576,
|
275 |
-
"eval_loss": 1.4818130731582642,
|
276 |
-
"eval_runtime": 1501.7616,
|
277 |
-
"eval_samples_per_second": 3.329,
|
278 |
-
"eval_steps_per_second": 0.416,
|
279 |
-
"step": 360
|
280 |
}
|
281 |
],
|
282 |
-
"logging_steps":
|
283 |
"max_steps": 1875,
|
284 |
"num_input_tokens_seen": 0,
|
285 |
"num_train_epochs": 3,
|
286 |
-
"save_steps":
|
287 |
"stateful_callbacks": {
|
288 |
"TrainerControl": {
|
289 |
"args": {
|
@@ -296,7 +41,7 @@
|
|
296 |
"attributes": {}
|
297 |
}
|
298 |
},
|
299 |
-
"total_flos":
|
300 |
"train_batch_size": 8,
|
301 |
"trial_name": null,
|
302 |
"trial_params": null
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.0816,
|
5 |
+
"eval_steps": 50,
|
6 |
+
"global_step": 51,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.08,
|
13 |
+
"grad_norm": 0.05632242560386658,
|
14 |
+
"learning_rate": 0.00013297872340425532,
|
15 |
+
"loss": 1.8852,
|
16 |
+
"step": 50
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 0.08,
|
20 |
+
"eval_loss": 1.7156720161437988,
|
21 |
+
"eval_runtime": 149.1859,
|
22 |
+
"eval_samples_per_second": 3.352,
|
23 |
+
"eval_steps_per_second": 0.422,
|
24 |
+
"step": 50
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25 |
}
|
26 |
],
|
27 |
+
"logging_steps": 50,
|
28 |
"max_steps": 1875,
|
29 |
"num_input_tokens_seen": 0,
|
30 |
"num_train_epochs": 3,
|
31 |
+
"save_steps": 50,
|
32 |
"stateful_callbacks": {
|
33 |
"TrainerControl": {
|
34 |
"args": {
|
|
|
41 |
"attributes": {}
|
42 |
}
|
43 |
},
|
44 |
+
"total_flos": 4002786222735360.0,
|
45 |
"train_batch_size": 8,
|
46 |
"trial_name": null,
|
47 |
"trial_params": null
|
tokenizer_config.json
CHANGED
@@ -121,11 +121,15 @@
|
|
121 |
"clean_up_tokenization_spaces": false,
|
122 |
"eos_token": "<|endoftext|>",
|
123 |
"legacy": false,
|
|
|
124 |
"model_max_length": 4096,
|
125 |
"pad_token": "<|endoftext|>",
|
126 |
"padding_side": "right",
|
127 |
"sp_model_kwargs": {},
|
|
|
128 |
"tokenizer_class": "LlamaTokenizer",
|
|
|
|
|
129 |
"unk_token": "<unk>",
|
130 |
"use_default_system_prompt": false
|
131 |
}
|
|
|
121 |
"clean_up_tokenization_spaces": false,
|
122 |
"eos_token": "<|endoftext|>",
|
123 |
"legacy": false,
|
124 |
+
"max_length": 512,
|
125 |
"model_max_length": 4096,
|
126 |
"pad_token": "<|endoftext|>",
|
127 |
"padding_side": "right",
|
128 |
"sp_model_kwargs": {},
|
129 |
+
"stride": 0,
|
130 |
"tokenizer_class": "LlamaTokenizer",
|
131 |
+
"truncation_side": "right",
|
132 |
+
"truncation_strategy": "longest_first",
|
133 |
"unk_token": "<unk>",
|
134 |
"use_default_system_prompt": false
|
135 |
}
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5432
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f6aadb86e7c11eaf53076fa580bc16d46538da6411436f54844013f1933386b8
|
3 |
size 5432
|