Hennara commited on
Commit
20e55d7
·
verified ·
1 Parent(s): 6e9cfe3

Add files using upload-large-folder tool

Browse files
Files changed (31) hide show
  1. S3r0-e4b/localizer/epoch-1/README.md +207 -0
  2. S3r0-e4b/localizer/epoch-1/adapter_config.json +43 -0
  3. S3r0-e4b/localizer/epoch-1/adapter_model.safetensors +3 -0
  4. S3r0-e4b/localizer/epoch-2/README.md +207 -0
  5. S3r0-e4b/localizer/epoch-2/adapter_config.json +43 -0
  6. S3r0-e4b/localizer/epoch-2/adapter_model.safetensors +3 -0
  7. S3r0-e4b/localizer/epoch-3/README.md +207 -0
  8. S3r0-e4b/localizer/epoch-3/adapter_config.json +43 -0
  9. S3r0-e4b/localizer/epoch-3/adapter_model.safetensors +3 -0
  10. S3r0-e4b/localizer/export/epoch-1/README.md +207 -0
  11. S3r0-e4b/localizer/export/epoch-1/adapter_config.json +51 -0
  12. S3r0-e4b/localizer/export/epoch-1/adapter_model.safetensors +3 -0
  13. S3r0-e4b/localizer/export/epoch-1/export.json +18 -0
  14. S3r0-e4b/localizer/export/epoch-2/README.md +207 -0
  15. S3r0-e4b/localizer/export/epoch-2/adapter_config.json +51 -0
  16. S3r0-e4b/localizer/export/epoch-2/adapter_model.safetensors +3 -0
  17. S3r0-e4b/localizer/export/epoch-2/export.json +18 -0
  18. S3r0-e4b/localizer/export/epoch-3/README.md +207 -0
  19. S3r0-e4b/localizer/export/epoch-3/adapter_config.json +51 -0
  20. S3r0-e4b/localizer/export/epoch-3/adapter_model.safetensors +3 -0
  21. S3r0-e4b/localizer/export/epoch-3/export.json +18 -0
  22. S3r0-e4b/localizer/export/final/README.md +207 -0
  23. S3r0-e4b/localizer/export/final/adapter_config.json +51 -0
  24. S3r0-e4b/localizer/export/final/adapter_model.safetensors +3 -0
  25. S3r0-e4b/localizer/export/final/export.json +18 -0
  26. S3r0-e4b/localizer/final/README.md +207 -0
  27. S3r0-e4b/localizer/final/adapter_config.json +43 -0
  28. S3r0-e4b/localizer/final/adapter_model.safetensors +3 -0
  29. S3r0-e4b/localizer/result.json +109 -0
  30. S3r0-e4b/localizer/run_meta.json +414 -0
  31. S3r0-e4b/localizer/val.jsonl +3 -0
S3r0-e4b/localizer/epoch-1/README.md ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: work/models/gemma4-e4b-w4a16-dequant-bf16
3
+ library_name: peft
4
+ pipeline_tag: text-generation
5
+ tags:
6
+ - base_model:adapter:work/models/gemma4-e4b-w4a16-dequant-bf16
7
+ - lora
8
+ - transformers
9
+ ---
10
+
11
+ # Model Card for Model ID
12
+
13
+ <!-- Provide a quick summary of what the model is/does. -->
14
+
15
+
16
+
17
+ ## Model Details
18
+
19
+ ### Model Description
20
+
21
+ <!-- Provide a longer summary of what this model is. -->
22
+
23
+
24
+
25
+ - **Developed by:** [More Information Needed]
26
+ - **Funded by [optional]:** [More Information Needed]
27
+ - **Shared by [optional]:** [More Information Needed]
28
+ - **Model type:** [More Information Needed]
29
+ - **Language(s) (NLP):** [More Information Needed]
30
+ - **License:** [More Information Needed]
31
+ - **Finetuned from model [optional]:** [More Information Needed]
32
+
33
+ ### Model Sources [optional]
34
+
35
+ <!-- Provide the basic links for the model. -->
36
+
37
+ - **Repository:** [More Information Needed]
38
+ - **Paper [optional]:** [More Information Needed]
39
+ - **Demo [optional]:** [More Information Needed]
40
+
41
+ ## Uses
42
+
43
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
44
+
45
+ ### Direct Use
46
+
47
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Downstream Use [optional]
52
+
53
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
54
+
55
+ [More Information Needed]
56
+
57
+ ### Out-of-Scope Use
58
+
59
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ## Bias, Risks, and Limitations
64
+
65
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
66
+
67
+ [More Information Needed]
68
+
69
+ ### Recommendations
70
+
71
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
72
+
73
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
74
+
75
+ ## How to Get Started with the Model
76
+
77
+ Use the code below to get started with the model.
78
+
79
+ [More Information Needed]
80
+
81
+ ## Training Details
82
+
83
+ ### Training Data
84
+
85
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
86
+
87
+ [More Information Needed]
88
+
89
+ ### Training Procedure
90
+
91
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
92
+
93
+ #### Preprocessing [optional]
94
+
95
+ [More Information Needed]
96
+
97
+
98
+ #### Training Hyperparameters
99
+
100
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
101
+
102
+ #### Speeds, Sizes, Times [optional]
103
+
104
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
105
+
106
+ [More Information Needed]
107
+
108
+ ## Evaluation
109
+
110
+ <!-- This section describes the evaluation protocols and provides the results. -->
111
+
112
+ ### Testing Data, Factors & Metrics
113
+
114
+ #### Testing Data
115
+
116
+ <!-- This should link to a Dataset Card if possible. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Factors
121
+
122
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
123
+
124
+ [More Information Needed]
125
+
126
+ #### Metrics
127
+
128
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
129
+
130
+ [More Information Needed]
131
+
132
+ ### Results
133
+
134
+ [More Information Needed]
135
+
136
+ #### Summary
137
+
138
+
139
+
140
+ ## Model Examination [optional]
141
+
142
+ <!-- Relevant interpretability work for the model goes here -->
143
+
144
+ [More Information Needed]
145
+
146
+ ## Environmental Impact
147
+
148
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
149
+
150
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
151
+
152
+ - **Hardware Type:** [More Information Needed]
153
+ - **Hours used:** [More Information Needed]
154
+ - **Cloud Provider:** [More Information Needed]
155
+ - **Compute Region:** [More Information Needed]
156
+ - **Carbon Emitted:** [More Information Needed]
157
+
158
+ ## Technical Specifications [optional]
159
+
160
+ ### Model Architecture and Objective
161
+
162
+ [More Information Needed]
163
+
164
+ ### Compute Infrastructure
165
+
166
+ [More Information Needed]
167
+
168
+ #### Hardware
169
+
170
+ [More Information Needed]
171
+
172
+ #### Software
173
+
174
+ [More Information Needed]
175
+
176
+ ## Citation [optional]
177
+
178
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
179
+
180
+ **BibTeX:**
181
+
182
+ [More Information Needed]
183
+
184
+ **APA:**
185
+
186
+ [More Information Needed]
187
+
188
+ ## Glossary [optional]
189
+
190
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
191
+
192
+ [More Information Needed]
193
+
194
+ ## More Information [optional]
195
+
196
+ [More Information Needed]
197
+
198
+ ## Model Card Authors [optional]
199
+
200
+ [More Information Needed]
201
+
202
+ ## Model Card Contact
203
+
204
+ [More Information Needed]
205
+ ### Framework versions
206
+
207
+ - PEFT 0.21.2
S3r0-e4b/localizer/epoch-1/adapter_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alora_invocation_tokens": null,
3
+ "alpha_pattern": {},
4
+ "arrow_config": null,
5
+ "auto_mapping": null,
6
+ "base_model_name_or_path": "work/models/gemma4-e4b-w4a16-dequant-bf16",
7
+ "bias": "none",
8
+ "corda_config": null,
9
+ "ensure_weight_tying": false,
10
+ "eva_config": null,
11
+ "exclude_modules": null,
12
+ "fan_in_fan_out": false,
13
+ "inference_mode": true,
14
+ "init_lora_weights": true,
15
+ "kasa_config": null,
16
+ "layer_replication": null,
17
+ "layers_pattern": null,
18
+ "layers_to_transform": null,
19
+ "loftq_config": {},
20
+ "lora_alpha": 32,
21
+ "lora_bias": false,
22
+ "lora_dropout": 0.0,
23
+ "lora_ga_config": null,
24
+ "megatron_config": null,
25
+ "megatron_core": "megatron.core",
26
+ "modules_to_save": null,
27
+ "monteclora_config": null,
28
+ "peft_type": "LORA",
29
+ "peft_version": "0.21.2",
30
+ "qalora_group_size": 16,
31
+ "r": 64,
32
+ "rank_pattern": {},
33
+ "revision": null,
34
+ "target_modules": "model\\.layers\\.\\d+\\.(self_attn|mlp)\\.(q_proj|k_proj|v_proj|o_proj|gate_proj|up_proj|down_proj)",
35
+ "target_parameters": null,
36
+ "task_type": "CAUSAL_LM",
37
+ "trainable_token_indices": null,
38
+ "use_bdlora": null,
39
+ "use_dora": false,
40
+ "use_qalora": false,
41
+ "use_rslora": false,
42
+ "velora_config": null
43
+ }
S3r0-e4b/localizer/epoch-1/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3bf37c453618e7cfb8837f11f9dae3d082d9f68b712dc2c98fe18e8dea791ac2
3
+ size 558174120
S3r0-e4b/localizer/epoch-2/README.md ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: work/models/gemma4-e4b-w4a16-dequant-bf16
3
+ library_name: peft
4
+ pipeline_tag: text-generation
5
+ tags:
6
+ - base_model:adapter:work/models/gemma4-e4b-w4a16-dequant-bf16
7
+ - lora
8
+ - transformers
9
+ ---
10
+
11
+ # Model Card for Model ID
12
+
13
+ <!-- Provide a quick summary of what the model is/does. -->
14
+
15
+
16
+
17
+ ## Model Details
18
+
19
+ ### Model Description
20
+
21
+ <!-- Provide a longer summary of what this model is. -->
22
+
23
+
24
+
25
+ - **Developed by:** [More Information Needed]
26
+ - **Funded by [optional]:** [More Information Needed]
27
+ - **Shared by [optional]:** [More Information Needed]
28
+ - **Model type:** [More Information Needed]
29
+ - **Language(s) (NLP):** [More Information Needed]
30
+ - **License:** [More Information Needed]
31
+ - **Finetuned from model [optional]:** [More Information Needed]
32
+
33
+ ### Model Sources [optional]
34
+
35
+ <!-- Provide the basic links for the model. -->
36
+
37
+ - **Repository:** [More Information Needed]
38
+ - **Paper [optional]:** [More Information Needed]
39
+ - **Demo [optional]:** [More Information Needed]
40
+
41
+ ## Uses
42
+
43
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
44
+
45
+ ### Direct Use
46
+
47
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Downstream Use [optional]
52
+
53
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
54
+
55
+ [More Information Needed]
56
+
57
+ ### Out-of-Scope Use
58
+
59
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ## Bias, Risks, and Limitations
64
+
65
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
66
+
67
+ [More Information Needed]
68
+
69
+ ### Recommendations
70
+
71
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
72
+
73
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
74
+
75
+ ## How to Get Started with the Model
76
+
77
+ Use the code below to get started with the model.
78
+
79
+ [More Information Needed]
80
+
81
+ ## Training Details
82
+
83
+ ### Training Data
84
+
85
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
86
+
87
+ [More Information Needed]
88
+
89
+ ### Training Procedure
90
+
91
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
92
+
93
+ #### Preprocessing [optional]
94
+
95
+ [More Information Needed]
96
+
97
+
98
+ #### Training Hyperparameters
99
+
100
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
101
+
102
+ #### Speeds, Sizes, Times [optional]
103
+
104
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
105
+
106
+ [More Information Needed]
107
+
108
+ ## Evaluation
109
+
110
+ <!-- This section describes the evaluation protocols and provides the results. -->
111
+
112
+ ### Testing Data, Factors & Metrics
113
+
114
+ #### Testing Data
115
+
116
+ <!-- This should link to a Dataset Card if possible. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Factors
121
+
122
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
123
+
124
+ [More Information Needed]
125
+
126
+ #### Metrics
127
+
128
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
129
+
130
+ [More Information Needed]
131
+
132
+ ### Results
133
+
134
+ [More Information Needed]
135
+
136
+ #### Summary
137
+
138
+
139
+
140
+ ## Model Examination [optional]
141
+
142
+ <!-- Relevant interpretability work for the model goes here -->
143
+
144
+ [More Information Needed]
145
+
146
+ ## Environmental Impact
147
+
148
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
149
+
150
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
151
+
152
+ - **Hardware Type:** [More Information Needed]
153
+ - **Hours used:** [More Information Needed]
154
+ - **Cloud Provider:** [More Information Needed]
155
+ - **Compute Region:** [More Information Needed]
156
+ - **Carbon Emitted:** [More Information Needed]
157
+
158
+ ## Technical Specifications [optional]
159
+
160
+ ### Model Architecture and Objective
161
+
162
+ [More Information Needed]
163
+
164
+ ### Compute Infrastructure
165
+
166
+ [More Information Needed]
167
+
168
+ #### Hardware
169
+
170
+ [More Information Needed]
171
+
172
+ #### Software
173
+
174
+ [More Information Needed]
175
+
176
+ ## Citation [optional]
177
+
178
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
179
+
180
+ **BibTeX:**
181
+
182
+ [More Information Needed]
183
+
184
+ **APA:**
185
+
186
+ [More Information Needed]
187
+
188
+ ## Glossary [optional]
189
+
190
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
191
+
192
+ [More Information Needed]
193
+
194
+ ## More Information [optional]
195
+
196
+ [More Information Needed]
197
+
198
+ ## Model Card Authors [optional]
199
+
200
+ [More Information Needed]
201
+
202
+ ## Model Card Contact
203
+
204
+ [More Information Needed]
205
+ ### Framework versions
206
+
207
+ - PEFT 0.21.2
S3r0-e4b/localizer/epoch-2/adapter_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alora_invocation_tokens": null,
3
+ "alpha_pattern": {},
4
+ "arrow_config": null,
5
+ "auto_mapping": null,
6
+ "base_model_name_or_path": "work/models/gemma4-e4b-w4a16-dequant-bf16",
7
+ "bias": "none",
8
+ "corda_config": null,
9
+ "ensure_weight_tying": false,
10
+ "eva_config": null,
11
+ "exclude_modules": null,
12
+ "fan_in_fan_out": false,
13
+ "inference_mode": true,
14
+ "init_lora_weights": true,
15
+ "kasa_config": null,
16
+ "layer_replication": null,
17
+ "layers_pattern": null,
18
+ "layers_to_transform": null,
19
+ "loftq_config": {},
20
+ "lora_alpha": 32,
21
+ "lora_bias": false,
22
+ "lora_dropout": 0.0,
23
+ "lora_ga_config": null,
24
+ "megatron_config": null,
25
+ "megatron_core": "megatron.core",
26
+ "modules_to_save": null,
27
+ "monteclora_config": null,
28
+ "peft_type": "LORA",
29
+ "peft_version": "0.21.2",
30
+ "qalora_group_size": 16,
31
+ "r": 64,
32
+ "rank_pattern": {},
33
+ "revision": null,
34
+ "target_modules": "model\\.layers\\.\\d+\\.(self_attn|mlp)\\.(q_proj|k_proj|v_proj|o_proj|gate_proj|up_proj|down_proj)",
35
+ "target_parameters": null,
36
+ "task_type": "CAUSAL_LM",
37
+ "trainable_token_indices": null,
38
+ "use_bdlora": null,
39
+ "use_dora": false,
40
+ "use_qalora": false,
41
+ "use_rslora": false,
42
+ "velora_config": null
43
+ }
S3r0-e4b/localizer/epoch-2/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f6e2a93da6bd15b50a35632a484fdff1e9db126c97b11b7c882aeafb0e80b981
3
+ size 558174120
S3r0-e4b/localizer/epoch-3/README.md ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: work/models/gemma4-e4b-w4a16-dequant-bf16
3
+ library_name: peft
4
+ pipeline_tag: text-generation
5
+ tags:
6
+ - base_model:adapter:work/models/gemma4-e4b-w4a16-dequant-bf16
7
+ - lora
8
+ - transformers
9
+ ---
10
+
11
+ # Model Card for Model ID
12
+
13
+ <!-- Provide a quick summary of what the model is/does. -->
14
+
15
+
16
+
17
+ ## Model Details
18
+
19
+ ### Model Description
20
+
21
+ <!-- Provide a longer summary of what this model is. -->
22
+
23
+
24
+
25
+ - **Developed by:** [More Information Needed]
26
+ - **Funded by [optional]:** [More Information Needed]
27
+ - **Shared by [optional]:** [More Information Needed]
28
+ - **Model type:** [More Information Needed]
29
+ - **Language(s) (NLP):** [More Information Needed]
30
+ - **License:** [More Information Needed]
31
+ - **Finetuned from model [optional]:** [More Information Needed]
32
+
33
+ ### Model Sources [optional]
34
+
35
+ <!-- Provide the basic links for the model. -->
36
+
37
+ - **Repository:** [More Information Needed]
38
+ - **Paper [optional]:** [More Information Needed]
39
+ - **Demo [optional]:** [More Information Needed]
40
+
41
+ ## Uses
42
+
43
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
44
+
45
+ ### Direct Use
46
+
47
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Downstream Use [optional]
52
+
53
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
54
+
55
+ [More Information Needed]
56
+
57
+ ### Out-of-Scope Use
58
+
59
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ## Bias, Risks, and Limitations
64
+
65
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
66
+
67
+ [More Information Needed]
68
+
69
+ ### Recommendations
70
+
71
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
72
+
73
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
74
+
75
+ ## How to Get Started with the Model
76
+
77
+ Use the code below to get started with the model.
78
+
79
+ [More Information Needed]
80
+
81
+ ## Training Details
82
+
83
+ ### Training Data
84
+
85
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
86
+
87
+ [More Information Needed]
88
+
89
+ ### Training Procedure
90
+
91
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
92
+
93
+ #### Preprocessing [optional]
94
+
95
+ [More Information Needed]
96
+
97
+
98
+ #### Training Hyperparameters
99
+
100
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
101
+
102
+ #### Speeds, Sizes, Times [optional]
103
+
104
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
105
+
106
+ [More Information Needed]
107
+
108
+ ## Evaluation
109
+
110
+ <!-- This section describes the evaluation protocols and provides the results. -->
111
+
112
+ ### Testing Data, Factors & Metrics
113
+
114
+ #### Testing Data
115
+
116
+ <!-- This should link to a Dataset Card if possible. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Factors
121
+
122
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
123
+
124
+ [More Information Needed]
125
+
126
+ #### Metrics
127
+
128
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
129
+
130
+ [More Information Needed]
131
+
132
+ ### Results
133
+
134
+ [More Information Needed]
135
+
136
+ #### Summary
137
+
138
+
139
+
140
+ ## Model Examination [optional]
141
+
142
+ <!-- Relevant interpretability work for the model goes here -->
143
+
144
+ [More Information Needed]
145
+
146
+ ## Environmental Impact
147
+
148
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
149
+
150
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
151
+
152
+ - **Hardware Type:** [More Information Needed]
153
+ - **Hours used:** [More Information Needed]
154
+ - **Cloud Provider:** [More Information Needed]
155
+ - **Compute Region:** [More Information Needed]
156
+ - **Carbon Emitted:** [More Information Needed]
157
+
158
+ ## Technical Specifications [optional]
159
+
160
+ ### Model Architecture and Objective
161
+
162
+ [More Information Needed]
163
+
164
+ ### Compute Infrastructure
165
+
166
+ [More Information Needed]
167
+
168
+ #### Hardware
169
+
170
+ [More Information Needed]
171
+
172
+ #### Software
173
+
174
+ [More Information Needed]
175
+
176
+ ## Citation [optional]
177
+
178
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
179
+
180
+ **BibTeX:**
181
+
182
+ [More Information Needed]
183
+
184
+ **APA:**
185
+
186
+ [More Information Needed]
187
+
188
+ ## Glossary [optional]
189
+
190
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
191
+
192
+ [More Information Needed]
193
+
194
+ ## More Information [optional]
195
+
196
+ [More Information Needed]
197
+
198
+ ## Model Card Authors [optional]
199
+
200
+ [More Information Needed]
201
+
202
+ ## Model Card Contact
203
+
204
+ [More Information Needed]
205
+ ### Framework versions
206
+
207
+ - PEFT 0.21.2
S3r0-e4b/localizer/epoch-3/adapter_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alora_invocation_tokens": null,
3
+ "alpha_pattern": {},
4
+ "arrow_config": null,
5
+ "auto_mapping": null,
6
+ "base_model_name_or_path": "work/models/gemma4-e4b-w4a16-dequant-bf16",
7
+ "bias": "none",
8
+ "corda_config": null,
9
+ "ensure_weight_tying": false,
10
+ "eva_config": null,
11
+ "exclude_modules": null,
12
+ "fan_in_fan_out": false,
13
+ "inference_mode": true,
14
+ "init_lora_weights": true,
15
+ "kasa_config": null,
16
+ "layer_replication": null,
17
+ "layers_pattern": null,
18
+ "layers_to_transform": null,
19
+ "loftq_config": {},
20
+ "lora_alpha": 32,
21
+ "lora_bias": false,
22
+ "lora_dropout": 0.0,
23
+ "lora_ga_config": null,
24
+ "megatron_config": null,
25
+ "megatron_core": "megatron.core",
26
+ "modules_to_save": null,
27
+ "monteclora_config": null,
28
+ "peft_type": "LORA",
29
+ "peft_version": "0.21.2",
30
+ "qalora_group_size": 16,
31
+ "r": 64,
32
+ "rank_pattern": {},
33
+ "revision": null,
34
+ "target_modules": "model\\.layers\\.\\d+\\.(self_attn|mlp)\\.(q_proj|k_proj|v_proj|o_proj|gate_proj|up_proj|down_proj)",
35
+ "target_parameters": null,
36
+ "task_type": "CAUSAL_LM",
37
+ "trainable_token_indices": null,
38
+ "use_bdlora": null,
39
+ "use_dora": false,
40
+ "use_qalora": false,
41
+ "use_rslora": false,
42
+ "velora_config": null
43
+ }
S3r0-e4b/localizer/epoch-3/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f9a2a2cc51c9fb91eb2f33501ca38fd2dbb750a229c17580698dbb7cb96508fd
3
+ size 558174120
S3r0-e4b/localizer/export/epoch-1/README.md ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: work/models/gemma4-e4b-w4a16-dequant-bf16
3
+ library_name: peft
4
+ pipeline_tag: text-generation
5
+ tags:
6
+ - base_model:adapter:work/models/gemma4-e4b-w4a16-dequant-bf16
7
+ - lora
8
+ - transformers
9
+ ---
10
+
11
+ # Model Card for Model ID
12
+
13
+ <!-- Provide a quick summary of what the model is/does. -->
14
+
15
+
16
+
17
+ ## Model Details
18
+
19
+ ### Model Description
20
+
21
+ <!-- Provide a longer summary of what this model is. -->
22
+
23
+
24
+
25
+ - **Developed by:** [More Information Needed]
26
+ - **Funded by [optional]:** [More Information Needed]
27
+ - **Shared by [optional]:** [More Information Needed]
28
+ - **Model type:** [More Information Needed]
29
+ - **Language(s) (NLP):** [More Information Needed]
30
+ - **License:** [More Information Needed]
31
+ - **Finetuned from model [optional]:** [More Information Needed]
32
+
33
+ ### Model Sources [optional]
34
+
35
+ <!-- Provide the basic links for the model. -->
36
+
37
+ - **Repository:** [More Information Needed]
38
+ - **Paper [optional]:** [More Information Needed]
39
+ - **Demo [optional]:** [More Information Needed]
40
+
41
+ ## Uses
42
+
43
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
44
+
45
+ ### Direct Use
46
+
47
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Downstream Use [optional]
52
+
53
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
54
+
55
+ [More Information Needed]
56
+
57
+ ### Out-of-Scope Use
58
+
59
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ## Bias, Risks, and Limitations
64
+
65
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
66
+
67
+ [More Information Needed]
68
+
69
+ ### Recommendations
70
+
71
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
72
+
73
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
74
+
75
+ ## How to Get Started with the Model
76
+
77
+ Use the code below to get started with the model.
78
+
79
+ [More Information Needed]
80
+
81
+ ## Training Details
82
+
83
+ ### Training Data
84
+
85
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
86
+
87
+ [More Information Needed]
88
+
89
+ ### Training Procedure
90
+
91
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
92
+
93
+ #### Preprocessing [optional]
94
+
95
+ [More Information Needed]
96
+
97
+
98
+ #### Training Hyperparameters
99
+
100
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
101
+
102
+ #### Speeds, Sizes, Times [optional]
103
+
104
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
105
+
106
+ [More Information Needed]
107
+
108
+ ## Evaluation
109
+
110
+ <!-- This section describes the evaluation protocols and provides the results. -->
111
+
112
+ ### Testing Data, Factors & Metrics
113
+
114
+ #### Testing Data
115
+
116
+ <!-- This should link to a Dataset Card if possible. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Factors
121
+
122
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
123
+
124
+ [More Information Needed]
125
+
126
+ #### Metrics
127
+
128
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
129
+
130
+ [More Information Needed]
131
+
132
+ ### Results
133
+
134
+ [More Information Needed]
135
+
136
+ #### Summary
137
+
138
+
139
+
140
+ ## Model Examination [optional]
141
+
142
+ <!-- Relevant interpretability work for the model goes here -->
143
+
144
+ [More Information Needed]
145
+
146
+ ## Environmental Impact
147
+
148
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
149
+
150
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
151
+
152
+ - **Hardware Type:** [More Information Needed]
153
+ - **Hours used:** [More Information Needed]
154
+ - **Cloud Provider:** [More Information Needed]
155
+ - **Compute Region:** [More Information Needed]
156
+ - **Carbon Emitted:** [More Information Needed]
157
+
158
+ ## Technical Specifications [optional]
159
+
160
+ ### Model Architecture and Objective
161
+
162
+ [More Information Needed]
163
+
164
+ ### Compute Infrastructure
165
+
166
+ [More Information Needed]
167
+
168
+ #### Hardware
169
+
170
+ [More Information Needed]
171
+
172
+ #### Software
173
+
174
+ [More Information Needed]
175
+
176
+ ## Citation [optional]
177
+
178
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
179
+
180
+ **BibTeX:**
181
+
182
+ [More Information Needed]
183
+
184
+ **APA:**
185
+
186
+ [More Information Needed]
187
+
188
+ ## Glossary [optional]
189
+
190
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
191
+
192
+ [More Information Needed]
193
+
194
+ ## More Information [optional]
195
+
196
+ [More Information Needed]
197
+
198
+ ## Model Card Authors [optional]
199
+
200
+ [More Information Needed]
201
+
202
+ ## Model Card Contact
203
+
204
+ [More Information Needed]
205
+ ### Framework versions
206
+
207
+ - PEFT 0.21.2
S3r0-e4b/localizer/export/epoch-1/adapter_config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alora_invocation_tokens": null,
3
+ "alpha_pattern": {},
4
+ "arrow_config": null,
5
+ "auto_mapping": null,
6
+ "base_model_name_or_path": "gemma-4-e4b-it-qat-w4a16-ct",
7
+ "bias": "none",
8
+ "corda_config": null,
9
+ "ensure_weight_tying": false,
10
+ "eva_config": null,
11
+ "exclude_modules": null,
12
+ "fan_in_fan_out": false,
13
+ "inference_mode": true,
14
+ "init_lora_weights": true,
15
+ "kasa_config": null,
16
+ "layer_replication": null,
17
+ "layers_pattern": null,
18
+ "layers_to_transform": null,
19
+ "loftq_config": {},
20
+ "lora_alpha": 32,
21
+ "lora_bias": false,
22
+ "lora_dropout": 0.0,
23
+ "lora_ga_config": null,
24
+ "megatron_config": null,
25
+ "megatron_core": "megatron.core",
26
+ "modules_to_save": null,
27
+ "monteclora_config": null,
28
+ "peft_type": "LORA",
29
+ "peft_version": "0.21.2",
30
+ "qalora_group_size": 16,
31
+ "r": 64,
32
+ "rank_pattern": {},
33
+ "revision": null,
34
+ "target_modules": [
35
+ "down_proj",
36
+ "gate_proj",
37
+ "k_proj",
38
+ "o_proj",
39
+ "q_proj",
40
+ "up_proj",
41
+ "v_proj"
42
+ ],
43
+ "target_parameters": null,
44
+ "task_type": "CAUSAL_LM",
45
+ "trainable_token_indices": null,
46
+ "use_bdlora": null,
47
+ "use_dora": false,
48
+ "use_qalora": false,
49
+ "use_rslora": false,
50
+ "velora_config": null
51
+ }
S3r0-e4b/localizer/export/epoch-1/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b404bc382ef09ba7e0bb5bc63a7d87d4d2b3fa6c45738e85de7e20b32769251
3
+ size 279122088
S3r0-e4b/localizer/export/epoch-1/export.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "r": 64,
3
+ "lora_alpha": 32,
4
+ "modules": 258,
5
+ "tensors": 516,
6
+ "k_eq_v_layers_copied": [],
7
+ "target_modules": [
8
+ "down_proj",
9
+ "gate_proj",
10
+ "k_proj",
11
+ "o_proj",
12
+ "q_proj",
13
+ "up_proj",
14
+ "v_proj"
15
+ ],
16
+ "bytes": 279128506,
17
+ "dtype": "bfloat16"
18
+ }
S3r0-e4b/localizer/export/epoch-2/README.md ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: work/models/gemma4-e4b-w4a16-dequant-bf16
3
+ library_name: peft
4
+ pipeline_tag: text-generation
5
+ tags:
6
+ - base_model:adapter:work/models/gemma4-e4b-w4a16-dequant-bf16
7
+ - lora
8
+ - transformers
9
+ ---
10
+
11
+ # Model Card for Model ID
12
+
13
+ <!-- Provide a quick summary of what the model is/does. -->
14
+
15
+
16
+
17
+ ## Model Details
18
+
19
+ ### Model Description
20
+
21
+ <!-- Provide a longer summary of what this model is. -->
22
+
23
+
24
+
25
+ - **Developed by:** [More Information Needed]
26
+ - **Funded by [optional]:** [More Information Needed]
27
+ - **Shared by [optional]:** [More Information Needed]
28
+ - **Model type:** [More Information Needed]
29
+ - **Language(s) (NLP):** [More Information Needed]
30
+ - **License:** [More Information Needed]
31
+ - **Finetuned from model [optional]:** [More Information Needed]
32
+
33
+ ### Model Sources [optional]
34
+
35
+ <!-- Provide the basic links for the model. -->
36
+
37
+ - **Repository:** [More Information Needed]
38
+ - **Paper [optional]:** [More Information Needed]
39
+ - **Demo [optional]:** [More Information Needed]
40
+
41
+ ## Uses
42
+
43
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
44
+
45
+ ### Direct Use
46
+
47
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Downstream Use [optional]
52
+
53
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
54
+
55
+ [More Information Needed]
56
+
57
+ ### Out-of-Scope Use
58
+
59
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ## Bias, Risks, and Limitations
64
+
65
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
66
+
67
+ [More Information Needed]
68
+
69
+ ### Recommendations
70
+
71
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
72
+
73
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
74
+
75
+ ## How to Get Started with the Model
76
+
77
+ Use the code below to get started with the model.
78
+
79
+ [More Information Needed]
80
+
81
+ ## Training Details
82
+
83
+ ### Training Data
84
+
85
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
86
+
87
+ [More Information Needed]
88
+
89
+ ### Training Procedure
90
+
91
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
92
+
93
+ #### Preprocessing [optional]
94
+
95
+ [More Information Needed]
96
+
97
+
98
+ #### Training Hyperparameters
99
+
100
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
101
+
102
+ #### Speeds, Sizes, Times [optional]
103
+
104
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
105
+
106
+ [More Information Needed]
107
+
108
+ ## Evaluation
109
+
110
+ <!-- This section describes the evaluation protocols and provides the results. -->
111
+
112
+ ### Testing Data, Factors & Metrics
113
+
114
+ #### Testing Data
115
+
116
+ <!-- This should link to a Dataset Card if possible. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Factors
121
+
122
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
123
+
124
+ [More Information Needed]
125
+
126
+ #### Metrics
127
+
128
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
129
+
130
+ [More Information Needed]
131
+
132
+ ### Results
133
+
134
+ [More Information Needed]
135
+
136
+ #### Summary
137
+
138
+
139
+
140
+ ## Model Examination [optional]
141
+
142
+ <!-- Relevant interpretability work for the model goes here -->
143
+
144
+ [More Information Needed]
145
+
146
+ ## Environmental Impact
147
+
148
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
149
+
150
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
151
+
152
+ - **Hardware Type:** [More Information Needed]
153
+ - **Hours used:** [More Information Needed]
154
+ - **Cloud Provider:** [More Information Needed]
155
+ - **Compute Region:** [More Information Needed]
156
+ - **Carbon Emitted:** [More Information Needed]
157
+
158
+ ## Technical Specifications [optional]
159
+
160
+ ### Model Architecture and Objective
161
+
162
+ [More Information Needed]
163
+
164
+ ### Compute Infrastructure
165
+
166
+ [More Information Needed]
167
+
168
+ #### Hardware
169
+
170
+ [More Information Needed]
171
+
172
+ #### Software
173
+
174
+ [More Information Needed]
175
+
176
+ ## Citation [optional]
177
+
178
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
179
+
180
+ **BibTeX:**
181
+
182
+ [More Information Needed]
183
+
184
+ **APA:**
185
+
186
+ [More Information Needed]
187
+
188
+ ## Glossary [optional]
189
+
190
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
191
+
192
+ [More Information Needed]
193
+
194
+ ## More Information [optional]
195
+
196
+ [More Information Needed]
197
+
198
+ ## Model Card Authors [optional]
199
+
200
+ [More Information Needed]
201
+
202
+ ## Model Card Contact
203
+
204
+ [More Information Needed]
205
+ ### Framework versions
206
+
207
+ - PEFT 0.21.2
S3r0-e4b/localizer/export/epoch-2/adapter_config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alora_invocation_tokens": null,
3
+ "alpha_pattern": {},
4
+ "arrow_config": null,
5
+ "auto_mapping": null,
6
+ "base_model_name_or_path": "gemma-4-e4b-it-qat-w4a16-ct",
7
+ "bias": "none",
8
+ "corda_config": null,
9
+ "ensure_weight_tying": false,
10
+ "eva_config": null,
11
+ "exclude_modules": null,
12
+ "fan_in_fan_out": false,
13
+ "inference_mode": true,
14
+ "init_lora_weights": true,
15
+ "kasa_config": null,
16
+ "layer_replication": null,
17
+ "layers_pattern": null,
18
+ "layers_to_transform": null,
19
+ "loftq_config": {},
20
+ "lora_alpha": 32,
21
+ "lora_bias": false,
22
+ "lora_dropout": 0.0,
23
+ "lora_ga_config": null,
24
+ "megatron_config": null,
25
+ "megatron_core": "megatron.core",
26
+ "modules_to_save": null,
27
+ "monteclora_config": null,
28
+ "peft_type": "LORA",
29
+ "peft_version": "0.21.2",
30
+ "qalora_group_size": 16,
31
+ "r": 64,
32
+ "rank_pattern": {},
33
+ "revision": null,
34
+ "target_modules": [
35
+ "down_proj",
36
+ "gate_proj",
37
+ "k_proj",
38
+ "o_proj",
39
+ "q_proj",
40
+ "up_proj",
41
+ "v_proj"
42
+ ],
43
+ "target_parameters": null,
44
+ "task_type": "CAUSAL_LM",
45
+ "trainable_token_indices": null,
46
+ "use_bdlora": null,
47
+ "use_dora": false,
48
+ "use_qalora": false,
49
+ "use_rslora": false,
50
+ "velora_config": null
51
+ }
S3r0-e4b/localizer/export/epoch-2/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:32513c2f1c0f10d0e4734cd38756981b1b7ad10bf66d4551777dfab0001c6fa7
3
+ size 279122088
S3r0-e4b/localizer/export/epoch-2/export.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "r": 64,
3
+ "lora_alpha": 32,
4
+ "modules": 258,
5
+ "tensors": 516,
6
+ "k_eq_v_layers_copied": [],
7
+ "target_modules": [
8
+ "down_proj",
9
+ "gate_proj",
10
+ "k_proj",
11
+ "o_proj",
12
+ "q_proj",
13
+ "up_proj",
14
+ "v_proj"
15
+ ],
16
+ "bytes": 279128506,
17
+ "dtype": "bfloat16"
18
+ }
S3r0-e4b/localizer/export/epoch-3/README.md ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: work/models/gemma4-e4b-w4a16-dequant-bf16
3
+ library_name: peft
4
+ pipeline_tag: text-generation
5
+ tags:
6
+ - base_model:adapter:work/models/gemma4-e4b-w4a16-dequant-bf16
7
+ - lora
8
+ - transformers
9
+ ---
10
+
11
+ # Model Card for Model ID
12
+
13
+ <!-- Provide a quick summary of what the model is/does. -->
14
+
15
+
16
+
17
+ ## Model Details
18
+
19
+ ### Model Description
20
+
21
+ <!-- Provide a longer summary of what this model is. -->
22
+
23
+
24
+
25
+ - **Developed by:** [More Information Needed]
26
+ - **Funded by [optional]:** [More Information Needed]
27
+ - **Shared by [optional]:** [More Information Needed]
28
+ - **Model type:** [More Information Needed]
29
+ - **Language(s) (NLP):** [More Information Needed]
30
+ - **License:** [More Information Needed]
31
+ - **Finetuned from model [optional]:** [More Information Needed]
32
+
33
+ ### Model Sources [optional]
34
+
35
+ <!-- Provide the basic links for the model. -->
36
+
37
+ - **Repository:** [More Information Needed]
38
+ - **Paper [optional]:** [More Information Needed]
39
+ - **Demo [optional]:** [More Information Needed]
40
+
41
+ ## Uses
42
+
43
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
44
+
45
+ ### Direct Use
46
+
47
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Downstream Use [optional]
52
+
53
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
54
+
55
+ [More Information Needed]
56
+
57
+ ### Out-of-Scope Use
58
+
59
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ## Bias, Risks, and Limitations
64
+
65
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
66
+
67
+ [More Information Needed]
68
+
69
+ ### Recommendations
70
+
71
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
72
+
73
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
74
+
75
+ ## How to Get Started with the Model
76
+
77
+ Use the code below to get started with the model.
78
+
79
+ [More Information Needed]
80
+
81
+ ## Training Details
82
+
83
+ ### Training Data
84
+
85
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
86
+
87
+ [More Information Needed]
88
+
89
+ ### Training Procedure
90
+
91
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
92
+
93
+ #### Preprocessing [optional]
94
+
95
+ [More Information Needed]
96
+
97
+
98
+ #### Training Hyperparameters
99
+
100
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
101
+
102
+ #### Speeds, Sizes, Times [optional]
103
+
104
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
105
+
106
+ [More Information Needed]
107
+
108
+ ## Evaluation
109
+
110
+ <!-- This section describes the evaluation protocols and provides the results. -->
111
+
112
+ ### Testing Data, Factors & Metrics
113
+
114
+ #### Testing Data
115
+
116
+ <!-- This should link to a Dataset Card if possible. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Factors
121
+
122
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
123
+
124
+ [More Information Needed]
125
+
126
+ #### Metrics
127
+
128
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
129
+
130
+ [More Information Needed]
131
+
132
+ ### Results
133
+
134
+ [More Information Needed]
135
+
136
+ #### Summary
137
+
138
+
139
+
140
+ ## Model Examination [optional]
141
+
142
+ <!-- Relevant interpretability work for the model goes here -->
143
+
144
+ [More Information Needed]
145
+
146
+ ## Environmental Impact
147
+
148
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
149
+
150
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
151
+
152
+ - **Hardware Type:** [More Information Needed]
153
+ - **Hours used:** [More Information Needed]
154
+ - **Cloud Provider:** [More Information Needed]
155
+ - **Compute Region:** [More Information Needed]
156
+ - **Carbon Emitted:** [More Information Needed]
157
+
158
+ ## Technical Specifications [optional]
159
+
160
+ ### Model Architecture and Objective
161
+
162
+ [More Information Needed]
163
+
164
+ ### Compute Infrastructure
165
+
166
+ [More Information Needed]
167
+
168
+ #### Hardware
169
+
170
+ [More Information Needed]
171
+
172
+ #### Software
173
+
174
+ [More Information Needed]
175
+
176
+ ## Citation [optional]
177
+
178
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
179
+
180
+ **BibTeX:**
181
+
182
+ [More Information Needed]
183
+
184
+ **APA:**
185
+
186
+ [More Information Needed]
187
+
188
+ ## Glossary [optional]
189
+
190
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
191
+
192
+ [More Information Needed]
193
+
194
+ ## More Information [optional]
195
+
196
+ [More Information Needed]
197
+
198
+ ## Model Card Authors [optional]
199
+
200
+ [More Information Needed]
201
+
202
+ ## Model Card Contact
203
+
204
+ [More Information Needed]
205
+ ### Framework versions
206
+
207
+ - PEFT 0.21.2
S3r0-e4b/localizer/export/epoch-3/adapter_config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alora_invocation_tokens": null,
3
+ "alpha_pattern": {},
4
+ "arrow_config": null,
5
+ "auto_mapping": null,
6
+ "base_model_name_or_path": "gemma-4-e4b-it-qat-w4a16-ct",
7
+ "bias": "none",
8
+ "corda_config": null,
9
+ "ensure_weight_tying": false,
10
+ "eva_config": null,
11
+ "exclude_modules": null,
12
+ "fan_in_fan_out": false,
13
+ "inference_mode": true,
14
+ "init_lora_weights": true,
15
+ "kasa_config": null,
16
+ "layer_replication": null,
17
+ "layers_pattern": null,
18
+ "layers_to_transform": null,
19
+ "loftq_config": {},
20
+ "lora_alpha": 32,
21
+ "lora_bias": false,
22
+ "lora_dropout": 0.0,
23
+ "lora_ga_config": null,
24
+ "megatron_config": null,
25
+ "megatron_core": "megatron.core",
26
+ "modules_to_save": null,
27
+ "monteclora_config": null,
28
+ "peft_type": "LORA",
29
+ "peft_version": "0.21.2",
30
+ "qalora_group_size": 16,
31
+ "r": 64,
32
+ "rank_pattern": {},
33
+ "revision": null,
34
+ "target_modules": [
35
+ "down_proj",
36
+ "gate_proj",
37
+ "k_proj",
38
+ "o_proj",
39
+ "q_proj",
40
+ "up_proj",
41
+ "v_proj"
42
+ ],
43
+ "target_parameters": null,
44
+ "task_type": "CAUSAL_LM",
45
+ "trainable_token_indices": null,
46
+ "use_bdlora": null,
47
+ "use_dora": false,
48
+ "use_qalora": false,
49
+ "use_rslora": false,
50
+ "velora_config": null
51
+ }
S3r0-e4b/localizer/export/epoch-3/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:500ca421af483524280a07d1ae93fe93810a8d95b5a047b38fff2784b53001db
3
+ size 279122088
S3r0-e4b/localizer/export/epoch-3/export.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "r": 64,
3
+ "lora_alpha": 32,
4
+ "modules": 258,
5
+ "tensors": 516,
6
+ "k_eq_v_layers_copied": [],
7
+ "target_modules": [
8
+ "down_proj",
9
+ "gate_proj",
10
+ "k_proj",
11
+ "o_proj",
12
+ "q_proj",
13
+ "up_proj",
14
+ "v_proj"
15
+ ],
16
+ "bytes": 279128506,
17
+ "dtype": "bfloat16"
18
+ }
S3r0-e4b/localizer/export/final/README.md ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: work/models/gemma4-e4b-w4a16-dequant-bf16
3
+ library_name: peft
4
+ pipeline_tag: text-generation
5
+ tags:
6
+ - base_model:adapter:work/models/gemma4-e4b-w4a16-dequant-bf16
7
+ - lora
8
+ - transformers
9
+ ---
10
+
11
+ # Model Card for Model ID
12
+
13
+ <!-- Provide a quick summary of what the model is/does. -->
14
+
15
+
16
+
17
+ ## Model Details
18
+
19
+ ### Model Description
20
+
21
+ <!-- Provide a longer summary of what this model is. -->
22
+
23
+
24
+
25
+ - **Developed by:** [More Information Needed]
26
+ - **Funded by [optional]:** [More Information Needed]
27
+ - **Shared by [optional]:** [More Information Needed]
28
+ - **Model type:** [More Information Needed]
29
+ - **Language(s) (NLP):** [More Information Needed]
30
+ - **License:** [More Information Needed]
31
+ - **Finetuned from model [optional]:** [More Information Needed]
32
+
33
+ ### Model Sources [optional]
34
+
35
+ <!-- Provide the basic links for the model. -->
36
+
37
+ - **Repository:** [More Information Needed]
38
+ - **Paper [optional]:** [More Information Needed]
39
+ - **Demo [optional]:** [More Information Needed]
40
+
41
+ ## Uses
42
+
43
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
44
+
45
+ ### Direct Use
46
+
47
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Downstream Use [optional]
52
+
53
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
54
+
55
+ [More Information Needed]
56
+
57
+ ### Out-of-Scope Use
58
+
59
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ## Bias, Risks, and Limitations
64
+
65
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
66
+
67
+ [More Information Needed]
68
+
69
+ ### Recommendations
70
+
71
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
72
+
73
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
74
+
75
+ ## How to Get Started with the Model
76
+
77
+ Use the code below to get started with the model.
78
+
79
+ [More Information Needed]
80
+
81
+ ## Training Details
82
+
83
+ ### Training Data
84
+
85
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
86
+
87
+ [More Information Needed]
88
+
89
+ ### Training Procedure
90
+
91
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
92
+
93
+ #### Preprocessing [optional]
94
+
95
+ [More Information Needed]
96
+
97
+
98
+ #### Training Hyperparameters
99
+
100
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
101
+
102
+ #### Speeds, Sizes, Times [optional]
103
+
104
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
105
+
106
+ [More Information Needed]
107
+
108
+ ## Evaluation
109
+
110
+ <!-- This section describes the evaluation protocols and provides the results. -->
111
+
112
+ ### Testing Data, Factors & Metrics
113
+
114
+ #### Testing Data
115
+
116
+ <!-- This should link to a Dataset Card if possible. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Factors
121
+
122
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
123
+
124
+ [More Information Needed]
125
+
126
+ #### Metrics
127
+
128
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
129
+
130
+ [More Information Needed]
131
+
132
+ ### Results
133
+
134
+ [More Information Needed]
135
+
136
+ #### Summary
137
+
138
+
139
+
140
+ ## Model Examination [optional]
141
+
142
+ <!-- Relevant interpretability work for the model goes here -->
143
+
144
+ [More Information Needed]
145
+
146
+ ## Environmental Impact
147
+
148
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
149
+
150
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
151
+
152
+ - **Hardware Type:** [More Information Needed]
153
+ - **Hours used:** [More Information Needed]
154
+ - **Cloud Provider:** [More Information Needed]
155
+ - **Compute Region:** [More Information Needed]
156
+ - **Carbon Emitted:** [More Information Needed]
157
+
158
+ ## Technical Specifications [optional]
159
+
160
+ ### Model Architecture and Objective
161
+
162
+ [More Information Needed]
163
+
164
+ ### Compute Infrastructure
165
+
166
+ [More Information Needed]
167
+
168
+ #### Hardware
169
+
170
+ [More Information Needed]
171
+
172
+ #### Software
173
+
174
+ [More Information Needed]
175
+
176
+ ## Citation [optional]
177
+
178
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
179
+
180
+ **BibTeX:**
181
+
182
+ [More Information Needed]
183
+
184
+ **APA:**
185
+
186
+ [More Information Needed]
187
+
188
+ ## Glossary [optional]
189
+
190
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
191
+
192
+ [More Information Needed]
193
+
194
+ ## More Information [optional]
195
+
196
+ [More Information Needed]
197
+
198
+ ## Model Card Authors [optional]
199
+
200
+ [More Information Needed]
201
+
202
+ ## Model Card Contact
203
+
204
+ [More Information Needed]
205
+ ### Framework versions
206
+
207
+ - PEFT 0.21.2
S3r0-e4b/localizer/export/final/adapter_config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alora_invocation_tokens": null,
3
+ "alpha_pattern": {},
4
+ "arrow_config": null,
5
+ "auto_mapping": null,
6
+ "base_model_name_or_path": "gemma-4-e4b-it-qat-w4a16-ct",
7
+ "bias": "none",
8
+ "corda_config": null,
9
+ "ensure_weight_tying": false,
10
+ "eva_config": null,
11
+ "exclude_modules": null,
12
+ "fan_in_fan_out": false,
13
+ "inference_mode": true,
14
+ "init_lora_weights": true,
15
+ "kasa_config": null,
16
+ "layer_replication": null,
17
+ "layers_pattern": null,
18
+ "layers_to_transform": null,
19
+ "loftq_config": {},
20
+ "lora_alpha": 32,
21
+ "lora_bias": false,
22
+ "lora_dropout": 0.0,
23
+ "lora_ga_config": null,
24
+ "megatron_config": null,
25
+ "megatron_core": "megatron.core",
26
+ "modules_to_save": null,
27
+ "monteclora_config": null,
28
+ "peft_type": "LORA",
29
+ "peft_version": "0.21.2",
30
+ "qalora_group_size": 16,
31
+ "r": 64,
32
+ "rank_pattern": {},
33
+ "revision": null,
34
+ "target_modules": [
35
+ "down_proj",
36
+ "gate_proj",
37
+ "k_proj",
38
+ "o_proj",
39
+ "q_proj",
40
+ "up_proj",
41
+ "v_proj"
42
+ ],
43
+ "target_parameters": null,
44
+ "task_type": "CAUSAL_LM",
45
+ "trainable_token_indices": null,
46
+ "use_bdlora": null,
47
+ "use_dora": false,
48
+ "use_qalora": false,
49
+ "use_rslora": false,
50
+ "velora_config": null
51
+ }
S3r0-e4b/localizer/export/final/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:500ca421af483524280a07d1ae93fe93810a8d95b5a047b38fff2784b53001db
3
+ size 279122088
S3r0-e4b/localizer/export/final/export.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "r": 64,
3
+ "lora_alpha": 32,
4
+ "modules": 258,
5
+ "tensors": 516,
6
+ "k_eq_v_layers_copied": [],
7
+ "target_modules": [
8
+ "down_proj",
9
+ "gate_proj",
10
+ "k_proj",
11
+ "o_proj",
12
+ "q_proj",
13
+ "up_proj",
14
+ "v_proj"
15
+ ],
16
+ "bytes": 279128506,
17
+ "dtype": "bfloat16"
18
+ }
S3r0-e4b/localizer/final/README.md ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: work/models/gemma4-e4b-w4a16-dequant-bf16
3
+ library_name: peft
4
+ pipeline_tag: text-generation
5
+ tags:
6
+ - base_model:adapter:work/models/gemma4-e4b-w4a16-dequant-bf16
7
+ - lora
8
+ - transformers
9
+ ---
10
+
11
+ # Model Card for Model ID
12
+
13
+ <!-- Provide a quick summary of what the model is/does. -->
14
+
15
+
16
+
17
+ ## Model Details
18
+
19
+ ### Model Description
20
+
21
+ <!-- Provide a longer summary of what this model is. -->
22
+
23
+
24
+
25
+ - **Developed by:** [More Information Needed]
26
+ - **Funded by [optional]:** [More Information Needed]
27
+ - **Shared by [optional]:** [More Information Needed]
28
+ - **Model type:** [More Information Needed]
29
+ - **Language(s) (NLP):** [More Information Needed]
30
+ - **License:** [More Information Needed]
31
+ - **Finetuned from model [optional]:** [More Information Needed]
32
+
33
+ ### Model Sources [optional]
34
+
35
+ <!-- Provide the basic links for the model. -->
36
+
37
+ - **Repository:** [More Information Needed]
38
+ - **Paper [optional]:** [More Information Needed]
39
+ - **Demo [optional]:** [More Information Needed]
40
+
41
+ ## Uses
42
+
43
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
44
+
45
+ ### Direct Use
46
+
47
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Downstream Use [optional]
52
+
53
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
54
+
55
+ [More Information Needed]
56
+
57
+ ### Out-of-Scope Use
58
+
59
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ## Bias, Risks, and Limitations
64
+
65
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
66
+
67
+ [More Information Needed]
68
+
69
+ ### Recommendations
70
+
71
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
72
+
73
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
74
+
75
+ ## How to Get Started with the Model
76
+
77
+ Use the code below to get started with the model.
78
+
79
+ [More Information Needed]
80
+
81
+ ## Training Details
82
+
83
+ ### Training Data
84
+
85
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
86
+
87
+ [More Information Needed]
88
+
89
+ ### Training Procedure
90
+
91
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
92
+
93
+ #### Preprocessing [optional]
94
+
95
+ [More Information Needed]
96
+
97
+
98
+ #### Training Hyperparameters
99
+
100
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
101
+
102
+ #### Speeds, Sizes, Times [optional]
103
+
104
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
105
+
106
+ [More Information Needed]
107
+
108
+ ## Evaluation
109
+
110
+ <!-- This section describes the evaluation protocols and provides the results. -->
111
+
112
+ ### Testing Data, Factors & Metrics
113
+
114
+ #### Testing Data
115
+
116
+ <!-- This should link to a Dataset Card if possible. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Factors
121
+
122
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
123
+
124
+ [More Information Needed]
125
+
126
+ #### Metrics
127
+
128
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
129
+
130
+ [More Information Needed]
131
+
132
+ ### Results
133
+
134
+ [More Information Needed]
135
+
136
+ #### Summary
137
+
138
+
139
+
140
+ ## Model Examination [optional]
141
+
142
+ <!-- Relevant interpretability work for the model goes here -->
143
+
144
+ [More Information Needed]
145
+
146
+ ## Environmental Impact
147
+
148
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
149
+
150
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
151
+
152
+ - **Hardware Type:** [More Information Needed]
153
+ - **Hours used:** [More Information Needed]
154
+ - **Cloud Provider:** [More Information Needed]
155
+ - **Compute Region:** [More Information Needed]
156
+ - **Carbon Emitted:** [More Information Needed]
157
+
158
+ ## Technical Specifications [optional]
159
+
160
+ ### Model Architecture and Objective
161
+
162
+ [More Information Needed]
163
+
164
+ ### Compute Infrastructure
165
+
166
+ [More Information Needed]
167
+
168
+ #### Hardware
169
+
170
+ [More Information Needed]
171
+
172
+ #### Software
173
+
174
+ [More Information Needed]
175
+
176
+ ## Citation [optional]
177
+
178
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
179
+
180
+ **BibTeX:**
181
+
182
+ [More Information Needed]
183
+
184
+ **APA:**
185
+
186
+ [More Information Needed]
187
+
188
+ ## Glossary [optional]
189
+
190
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
191
+
192
+ [More Information Needed]
193
+
194
+ ## More Information [optional]
195
+
196
+ [More Information Needed]
197
+
198
+ ## Model Card Authors [optional]
199
+
200
+ [More Information Needed]
201
+
202
+ ## Model Card Contact
203
+
204
+ [More Information Needed]
205
+ ### Framework versions
206
+
207
+ - PEFT 0.21.2
S3r0-e4b/localizer/final/adapter_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alora_invocation_tokens": null,
3
+ "alpha_pattern": {},
4
+ "arrow_config": null,
5
+ "auto_mapping": null,
6
+ "base_model_name_or_path": "work/models/gemma4-e4b-w4a16-dequant-bf16",
7
+ "bias": "none",
8
+ "corda_config": null,
9
+ "ensure_weight_tying": false,
10
+ "eva_config": null,
11
+ "exclude_modules": null,
12
+ "fan_in_fan_out": false,
13
+ "inference_mode": true,
14
+ "init_lora_weights": true,
15
+ "kasa_config": null,
16
+ "layer_replication": null,
17
+ "layers_pattern": null,
18
+ "layers_to_transform": null,
19
+ "loftq_config": {},
20
+ "lora_alpha": 32,
21
+ "lora_bias": false,
22
+ "lora_dropout": 0.0,
23
+ "lora_ga_config": null,
24
+ "megatron_config": null,
25
+ "megatron_core": "megatron.core",
26
+ "modules_to_save": null,
27
+ "monteclora_config": null,
28
+ "peft_type": "LORA",
29
+ "peft_version": "0.21.2",
30
+ "qalora_group_size": 16,
31
+ "r": 64,
32
+ "rank_pattern": {},
33
+ "revision": null,
34
+ "target_modules": "model\\.layers\\.\\d+\\.(self_attn|mlp)\\.(q_proj|k_proj|v_proj|o_proj|gate_proj|up_proj|down_proj)",
35
+ "target_parameters": null,
36
+ "task_type": "CAUSAL_LM",
37
+ "trainable_token_indices": null,
38
+ "use_bdlora": null,
39
+ "use_dora": false,
40
+ "use_qalora": false,
41
+ "use_rslora": false,
42
+ "velora_config": null
43
+ }
S3r0-e4b/localizer/final/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f9a2a2cc51c9fb91eb2f33501ca38fd2dbb750a229c17580698dbb7cb96508fd
3
+ size 558174120
S3r0-e4b/localizer/result.json ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": {
3
+ "train_runtime": 4355.8559,
4
+ "train_samples_per_second": 0.871,
5
+ "train_steps_per_second": 0.054,
6
+ "total_flos": 0.0,
7
+ "train_loss": 0.6857871996073783,
8
+ "epoch": 1.0
9
+ },
10
+ "plan": {
11
+ "optimizer_steps": 237,
12
+ "rows": 3792,
13
+ "rows_per_step": 16,
14
+ "dropped_tail_rows": 0,
15
+ "examples_used": 3792,
16
+ "examples_per_source": {
17
+ "s3_loc_r0": 3792
18
+ },
19
+ "tokens": 24493758,
20
+ "trained_tokens": 4206273,
21
+ "mean_row_tokens": 6459,
22
+ "tasks": 1264,
23
+ "visits_per_task": {
24
+ "min": 3,
25
+ "median": 3,
26
+ "max": 3
27
+ },
28
+ "epoch_end_steps": [
29
+ 79,
30
+ 158,
31
+ 237
32
+ ]
33
+ },
34
+ "final_adapter": "work/runs/S3r0-e4b/localizer/final",
35
+ "exports": {
36
+ "epoch-1": {
37
+ "r": 64,
38
+ "lora_alpha": 32,
39
+ "modules": 258,
40
+ "tensors": 516,
41
+ "k_eq_v_layers_copied": [],
42
+ "target_modules": [
43
+ "down_proj",
44
+ "gate_proj",
45
+ "k_proj",
46
+ "o_proj",
47
+ "q_proj",
48
+ "up_proj",
49
+ "v_proj"
50
+ ],
51
+ "bytes": 279128506,
52
+ "dtype": "bfloat16"
53
+ },
54
+ "epoch-2": {
55
+ "r": 64,
56
+ "lora_alpha": 32,
57
+ "modules": 258,
58
+ "tensors": 516,
59
+ "k_eq_v_layers_copied": [],
60
+ "target_modules": [
61
+ "down_proj",
62
+ "gate_proj",
63
+ "k_proj",
64
+ "o_proj",
65
+ "q_proj",
66
+ "up_proj",
67
+ "v_proj"
68
+ ],
69
+ "bytes": 279128506,
70
+ "dtype": "bfloat16"
71
+ },
72
+ "epoch-3": {
73
+ "r": 64,
74
+ "lora_alpha": 32,
75
+ "modules": 258,
76
+ "tensors": 516,
77
+ "k_eq_v_layers_copied": [],
78
+ "target_modules": [
79
+ "down_proj",
80
+ "gate_proj",
81
+ "k_proj",
82
+ "o_proj",
83
+ "q_proj",
84
+ "up_proj",
85
+ "v_proj"
86
+ ],
87
+ "bytes": 279128506,
88
+ "dtype": "bfloat16"
89
+ },
90
+ "final": {
91
+ "r": 64,
92
+ "lora_alpha": 32,
93
+ "modules": 258,
94
+ "tensors": 516,
95
+ "k_eq_v_layers_copied": [],
96
+ "target_modules": [
97
+ "down_proj",
98
+ "gate_proj",
99
+ "k_proj",
100
+ "o_proj",
101
+ "q_proj",
102
+ "up_proj",
103
+ "v_proj"
104
+ ],
105
+ "bytes": 279128506,
106
+ "dtype": "bfloat16"
107
+ }
108
+ }
109
+ }
S3r0-e4b/localizer/run_meta.json ADDED
@@ -0,0 +1,414 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "created": "2026-11-05T07:07:54.737835+00:00",
3
+ "argv": [
4
+ "/workspace/locfix/src/locfix/cli.py",
5
+ "train",
6
+ "--config",
7
+ "configs/runs/S3r0-e4b-localizer.yaml",
8
+ "--set",
9
+ "train.base_model=work/models/gemma4-e4b-w4a16-dequant-bf16",
10
+ "--set",
11
+ "train.learning_rate=2e-4",
12
+ "--set",
13
+ "train.global_batch_sequences=16",
14
+ "--set",
15
+ "train.per_device_batch=1",
16
+ "--set",
17
+ "train.gradient_accumulation_steps=4",
18
+ "--set",
19
+ "train.schedule.warmup_steps=10",
20
+ "--set",
21
+ "train.schedule.min_lr_ratio=0.1",
22
+ "--set",
23
+ "train.max_seq_len=32768",
24
+ "--set",
25
+ "train.lora.r=64",
26
+ "--set",
27
+ "train.lora.alpha=32",
28
+ "--set",
29
+ "train.lora.dropout=0.0",
30
+ "--set",
31
+ "train.optimizer.weight_decay=0.0",
32
+ "--set",
33
+ "train.optimizer.grad_clip=1.0",
34
+ "--set",
35
+ "train.packing=false",
36
+ "--set",
37
+ "train.length_grouping=true",
38
+ "--set",
39
+ "seed=1234",
40
+ "--set",
41
+ "train.output_dir=work/runs/S3r0-e4b/localizer",
42
+ "--set",
43
+ "train.epochs=3"
44
+ ],
45
+ "git_sha": "2a4d47a6d6c7f0f283620ee62743acdd973df906-dirty",
46
+ "python": "3.12.3",
47
+ "versions": {
48
+ "locfix": "0.1.0",
49
+ "swegemma": null,
50
+ "adk-submission": null,
51
+ "adk-eval-core": null,
52
+ "google-adk": null,
53
+ "litellm": null,
54
+ "transformers": "5.13.1",
55
+ "peft": "0.21.2",
56
+ "trl": null,
57
+ "accelerate": "1.15.0",
58
+ "torch": "2.10.0",
59
+ "vllm": "0.19.1",
60
+ "liger-kernel": null
61
+ },
62
+ "config": {
63
+ "seed": 1234,
64
+ "paths": {
65
+ "data_root": "./work/data/codeagent_final",
66
+ "competition_data": "./work/data/gemma4-dev",
67
+ "vendor_wheels": "./vendor",
68
+ "models": "./work/models",
69
+ "records": "./work/records",
70
+ "runs": "./work/runs",
71
+ "eval": "./work/eval",
72
+ "golden": "./work/golden",
73
+ "codeagent_validate": "./work/codeagent-validate"
74
+ },
75
+ "versions": {
76
+ "vllm": "0.19.1",
77
+ "transformers": "5.13.1",
78
+ "swegemma": "0.2.11",
79
+ "adk_submission": "0.2.13",
80
+ "adk_eval_core": "0.1.0"
81
+ },
82
+ "models": {
83
+ "e4b": {
84
+ "serve_id": "gemma-4-e4b-it-qat-w4a16-ct",
85
+ "hf_ct": "google/gemma-4-e4b-it-qat-w4a16-ct",
86
+ "kaggle": "google/gemma-4/other/gemma-4-e4b-it-qat-w4a16-ct/2",
87
+ "dequantized": "./work/models/gemma4-e4b-w4a16-dequant-bf16",
88
+ "lora_targets": [
89
+ "q_proj",
90
+ "k_proj",
91
+ "v_proj",
92
+ "o_proj",
93
+ "gate_proj",
94
+ "up_proj",
95
+ "down_proj"
96
+ ]
97
+ },
98
+ "31b": {
99
+ "serve_id": "gemma-4-31b-it-qat-w4a16-ct",
100
+ "hf_ct": "google/gemma-4-31b-it-qat-w4a16-ct",
101
+ "kaggle": "google/gemma-4/other/gemma-4-31b-it-qat-w4a16-ct/2",
102
+ "dequantized": "./work/models/gemma4-31b-w4a16-dequant-bf16",
103
+ "lora_targets": [
104
+ "q_proj",
105
+ "k_proj",
106
+ "v_proj",
107
+ "o_proj",
108
+ "gate_proj",
109
+ "up_proj",
110
+ "down_proj"
111
+ ]
112
+ },
113
+ "12b": {
114
+ "serve_id": "gemma-4-12b-it-qat-w4a16-ct",
115
+ "hf_ct": "google/gemma-4-12b-it-qat-w4a16-ct",
116
+ "dequantized": "./work/models/gemma4-12b-w4a16-dequant-bf16",
117
+ "text_only_class": "Gemma4ForCausalLM",
118
+ "lora_targets": [
119
+ "q_proj",
120
+ "k_proj",
121
+ "v_proj",
122
+ "o_proj",
123
+ "gate_proj",
124
+ "up_proj",
125
+ "down_proj"
126
+ ]
127
+ }
128
+ },
129
+ "harness": {
130
+ "max_model_len": 32768,
131
+ "tool_call_parser": "gemma4",
132
+ "reasoning_parser": "gemma4",
133
+ "compaction": {
134
+ "compaction_interval": 15,
135
+ "overlap_size": 2,
136
+ "token_threshold": 14336,
137
+ "event_retention_size": 5
138
+ },
139
+ "context_cache": {
140
+ "min_tokens": 2048,
141
+ "ttl_seconds": 1800,
142
+ "cache_intervals": 10
143
+ },
144
+ "read_file_max_lines": 150,
145
+ "read_file_max_chars": 10000,
146
+ "max_stdout_chars": 5000,
147
+ "command_timeout_seconds": 300
148
+ },
149
+ "budget": {
150
+ "max_time_minutes": 8,
151
+ "max_tool_calls": 36,
152
+ "max_turns": 300
153
+ },
154
+ "budget_locfix": {
155
+ "max_time_minutes": 8,
156
+ "max_tool_calls": 44,
157
+ "max_turns": 300
158
+ },
159
+ "timing": {
160
+ "e4b": {
161
+ "prefill_tps": null,
162
+ "decode_tps": null,
163
+ "setup_seconds": null
164
+ },
165
+ "31b": {
166
+ "prefill_tps": null,
167
+ "decode_tps": null,
168
+ "setup_seconds": null
169
+ }
170
+ },
171
+ "generation": {
172
+ "include_thoughts": true,
173
+ "thinking_budget": 4096,
174
+ "max_output_tokens": 16384
175
+ },
176
+ "data": {
177
+ "training_layout": "auto",
178
+ "replay_compaction": false,
179
+ "max_trace_tokens": 32768,
180
+ "teacher_time_minutes": 8,
181
+ "time_scale": "auto",
182
+ "val_fraction": 0.03,
183
+ "exclude_repos": [
184
+ "fastapi",
185
+ "requests",
186
+ "rich",
187
+ "httpx"
188
+ ],
189
+ "heldout_size": 300
190
+ },
191
+ "train": {
192
+ "model": "e4b",
193
+ "role": "localizer",
194
+ "sources": [
195
+ {
196
+ "path": "./work/records/locfix/s3_loc_r0.jsonl"
197
+ }
198
+ ],
199
+ "val_sources": [
200
+ {
201
+ "path": "./work/records/val/locfix/s3_loc_r0.jsonl"
202
+ }
203
+ ],
204
+ "output_dir": "work/runs/S3r0-e4b/localizer",
205
+ "base_model": "work/models/gemma4-e4b-w4a16-dequant-bf16",
206
+ "tokenizer": null,
207
+ "init_adapter": null,
208
+ "lr_scale": 1.0,
209
+ "learning_rate": 0.0002,
210
+ "max_seq_len": 32768,
211
+ "lora": {
212
+ "r": 64,
213
+ "alpha": 32,
214
+ "dropout": 0.0,
215
+ "use_rslora": false
216
+ },
217
+ "optimizer": {
218
+ "name": "adamw",
219
+ "hf_name": "adamw_torch",
220
+ "betas": [
221
+ 0.9,
222
+ 0.999
223
+ ],
224
+ "weight_decay": 0.0,
225
+ "grad_clip": 1.0
226
+ },
227
+ "schedule": {
228
+ "warmup_steps": 10,
229
+ "decay": "cosine",
230
+ "min_lr_ratio": 0.1
231
+ },
232
+ "global_batch_sequences": 16,
233
+ "per_device_batch": 1,
234
+ "gradient_accumulation_steps": 4,
235
+ "gradient_checkpointing": true,
236
+ "packing": false,
237
+ "length_grouping": true,
238
+ "attn_implementation": "sdpa",
239
+ "ce_chunk": 1024,
240
+ "bf16": true,
241
+ "cache_dir": "./work/records/token_cache",
242
+ "logging_steps": 1,
243
+ "report_to": [],
244
+ "export": true,
245
+ "lr": {
246
+ "e4b": 0.0002,
247
+ "31b": 0.0001,
248
+ "12b": 0.0001
249
+ },
250
+ "epochs": 3,
251
+ "save_every_epoch": true,
252
+ "warm_start_lr_scale": 0.5,
253
+ "target_visits_per_patch": 3
254
+ },
255
+ "search": {
256
+ "lr": {
257
+ "e4b": [
258
+ 0.0001,
259
+ 0.0002,
260
+ 0.0004
261
+ ],
262
+ "31b": [
263
+ 5e-05,
264
+ 0.0001,
265
+ 0.0002
266
+ ]
267
+ },
268
+ "rank": {
269
+ "e4b": [
270
+ 32,
271
+ 64
272
+ ],
273
+ "31b": [
274
+ 32,
275
+ 64
276
+ ]
277
+ },
278
+ "global_batch_sequences": [
279
+ 8,
280
+ 16
281
+ ]
282
+ },
283
+ "em": {
284
+ "K": 8,
285
+ "sample_temperature": 0.8,
286
+ "ess_target": 2,
287
+ "ctx_cost_sigma": 0.25,
288
+ "req_cost_sigma": 0.25,
289
+ "beta": 0.5,
290
+ "resample_m": 3,
291
+ "no_request_fraction": 0.4,
292
+ "max_locations": 8,
293
+ "ctx_budget_tokens": 6000,
294
+ "localizer_max_calls": 15,
295
+ "max_rounds": 3,
296
+ "unsolved_ratio": 0.0
297
+ },
298
+ "eval": {
299
+ "seeds": [
300
+ 0,
301
+ 1
302
+ ],
303
+ "concurrency": 32
304
+ }
305
+ },
306
+ "settings": {
307
+ "output_dir": "work/runs/S3r0-e4b/localizer",
308
+ "base_model": "work/models/gemma4-e4b-w4a16-dequant-bf16",
309
+ "sources": [
310
+ {
311
+ "path": "./work/records/locfix/s3_loc_r0.jsonl",
312
+ "weight": 1.0,
313
+ "repeat": 1.0,
314
+ "name": null
315
+ }
316
+ ],
317
+ "val_sources": [
318
+ {
319
+ "path": "./work/records/val/locfix/s3_loc_r0.jsonl",
320
+ "weight": 1.0,
321
+ "repeat": 1.0,
322
+ "name": null
323
+ }
324
+ ],
325
+ "tokenizer": null,
326
+ "role": "localizer",
327
+ "model_key": "e4b",
328
+ "epochs": 3,
329
+ "learning_rate": 0.0002,
330
+ "lora_r": 64,
331
+ "lora_alpha": 32,
332
+ "lora_dropout": 0.0,
333
+ "init_adapter": null,
334
+ "global_batch_rows": 16,
335
+ "per_device_rows": 1,
336
+ "gradient_accumulation": 4,
337
+ "max_seq_len": 32768,
338
+ "packing": false,
339
+ "length_grouping": true,
340
+ "attn_implementation": "sdpa",
341
+ "gradient_checkpointing": true,
342
+ "ce_chunk": 1024,
343
+ "warmup_steps": 10,
344
+ "min_lr_ratio": 0.1,
345
+ "weight_decay": 0.0,
346
+ "betas": [
347
+ 0.9,
348
+ 0.999
349
+ ],
350
+ "grad_clip": 1.0,
351
+ "seed": 1234,
352
+ "cache_dir": "./work/records/token_cache",
353
+ "bf16": true,
354
+ "logging_steps": 1,
355
+ "report_to": [],
356
+ "export": true,
357
+ "serve_id": "gemma-4-e4b-it-qat-w4a16-ct",
358
+ "optim": "adamw_torch",
359
+ "lr_scheduler": "cosine_with_min_lr"
360
+ },
361
+ "world_size": 4,
362
+ "gradient_accumulation_steps": 4,
363
+ "plan": {
364
+ "optimizer_steps": 237,
365
+ "rows": 3792,
366
+ "rows_per_step": 16,
367
+ "dropped_tail_rows": 0,
368
+ "examples_used": 3792,
369
+ "examples_per_source": {
370
+ "s3_loc_r0": 3792
371
+ },
372
+ "tokens": 24493758,
373
+ "trained_tokens": 4206273,
374
+ "mean_row_tokens": 6459,
375
+ "tasks": 1264,
376
+ "visits_per_task": {
377
+ "min": 3,
378
+ "median": 3,
379
+ "max": 3
380
+ },
381
+ "epoch_end_steps": [
382
+ 79,
383
+ 158,
384
+ 237
385
+ ]
386
+ },
387
+ "sources": [
388
+ {
389
+ "source": "work/records/locfix/s3_loc_r0.jsonl",
390
+ "records": 1264,
391
+ "examples": 1264,
392
+ "too_long": 0,
393
+ "render_errors": 0,
394
+ "no_trained_tokens": 0,
395
+ "tokens": 8164586,
396
+ "trained_tokens": 1402091,
397
+ "errors": []
398
+ }
399
+ ],
400
+ "val_sources": [
401
+ {
402
+ "source": "work/records/val/locfix/s3_loc_r0.jsonl",
403
+ "records": 39,
404
+ "examples": 39,
405
+ "too_long": 0,
406
+ "render_errors": 0,
407
+ "no_trained_tokens": 0,
408
+ "tokens": 233365,
409
+ "trained_tokens": 36550,
410
+ "errors": []
411
+ }
412
+ ],
413
+ "trainable_parameters": 139526144
414
+ }
S3r0-e4b/localizer/val.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {"epoch": 1, "step": 79, "val_nll/s3_loc_r0": 0.75915, "val_nll/all": 0.75915}
2
+ {"epoch": 2, "step": 158, "val_nll/s3_loc_r0": 0.72056, "val_nll/all": 0.72056}
3
+ {"epoch": 3, "step": 237, "val_nll/s3_loc_r0": 0.71995, "val_nll/all": 0.71995}