Commits · inflaton/machine-translation

v2e results

83aa28d

dh-mc commited on Oct 13

Update 03c_RAPGeT_v2_Data Analysis.ipynb

9ad0cbb

dh-mc commited on Oct 12

Chinese-English Machine Translation - Repetition Metrics for Different Models

a5c9876

dh-mc commited on Oct 12

analyzed llama-3-8b rpp1.0

0f06270

dh-mc commited on Oct 12

Update 03b_RAPGeT_v2_Data Analysis_Generic_Prompt.ipynb

d523de0

dh-mc commited on Oct 11

completed llama-3.1-70b generic prompt

db47db5

dh-mc commited on Oct 11

analyzed llama-3.1-70b rpp=1

6d425fc

dh-mc commited on Oct 6

added MTR graphs

f62fd81

dh-mc commited on Oct 6

ready for llama-3.1-8b

b83b354

dh-mc commited on Oct 6

chat template COMET

1ebd37e

inflaton commited on Oct 6

COMET eval results

927a69c

dh-mc commited on Sep 17

few shot COMET results

101f384

dh-mc commited on Sep 16

eval comet with rtx4080

243d523

inflaton commited on Sep 15

diable lfs for 00b notebook

553af0d

dh-mc commited on Sep 15

Update 00c_Data Analysis_Fine_Tuned.ipynb

80027ea

dh-mc commited on Aug 29

fine-tuned gpt-4o

46d9e0e

dh-mc commited on Aug 29

completed results: max_new_tokens = 2048 and new addition of Phi-3.5-mini (3.8B)

a35e6bd

dh-mc commited on Aug 26

ready to eval phi-3.5

25ab915

dh-mc commited on Aug 25

refine graph

b04757c

dh-mc commited on Aug 22

completed RPP analysis

0cc187e

dh-mc commited on Aug 17

completed analysis for llama-3.1-70b

bdc91f5

dh-mc commited on Aug 13

complete gpt-4o-mini training

3f6b774

dh-mc commited on Aug 10

add env var END_REPETITION_PENALTY

ee71b10

dh-mc commited on Aug 9

fine-tuned metrics WIP

2a89293

inflaton commited on Aug 9

train with 4gpu

a2100ac

dh-mc commited on Aug 8

ready for LF lora training

6fc6dc9

dh-mc commited on Aug 7

translation time analyzed

95ba169

dh-mc commited on Aug 6

complete results for gpt-4o and gpt-4o-mini

fddc7fb

inflaton commited on Aug 5

Update 00b_Data Analysis_Few_Shots.ipynb

60b04fa

dh-mc commited on Aug 4

refined data analysis

9a2b128

dh-mc commited on Aug 4

complete results for original prompt

ad9aba8

dh-mc commited on Aug 4

ready for few shots prompting 4gpu

0156aec

dh-mc commited on Aug 3

eval with few-shots prompting

6c91c84

dh-mc commited on Aug 3

enable env: START_REPETITION_PENALTY

a457091

dh-mc commited on Aug 2

completed Qwen2-72B results

7dab5c1

dh-mc commited on Aug 2

new mistral results analyzed

1bdff66

dh-mc commited on Jul 31

set max_new_tokens to 300

a37d279

dh-mc commited on Jul 30

formatted code

640a737

dh-mc commited on Jul 30

data analysis notebooks

1a9edc9

dh-mc commited on Jul 30

fixed bug for Mistral

12a5ff3

dh-mc commited on Jul 30

enable do_sample

a69b127

dh-mc commited on Jul 30

set max_new_tokens to 2048

990ef6d

dh-mc commited on Jul 29

ready for gpu cluster

c73d190

dh-mc commited on Jul 28

initial code for Chinese/English translation

3860729

dh-mc commited on Jul 15

Commit History

v2e results 83aa28d

Update 03c_RAPGeT_v2_Data Analysis.ipynb 9ad0cbb

Chinese-English Machine Translation - Repetition Metrics for Different Models a5c9876

analyzed llama-3-8b rpp1.0 0f06270

Update 03b_RAPGeT_v2_Data Analysis_Generic_Prompt.ipynb d523de0

completed llama-3.1-70b generic prompt db47db5

analyzed llama-3.1-70b rpp=1 6d425fc

added MTR graphs f62fd81

ready for llama-3.1-8b b83b354

chat template COMET 1ebd37e

COMET eval results 927a69c

few shot COMET results 101f384

eval comet with rtx4080 243d523

diable lfs for 00b notebook 553af0d

Update 00c_Data Analysis_Fine_Tuned.ipynb 80027ea

fine-tuned gpt-4o 46d9e0e

completed results: max_new_tokens = 2048 and new addition of Phi-3.5-mini (3.8B) a35e6bd

ready to eval phi-3.5 25ab915

refine graph b04757c

completed RPP analysis 0cc187e

completed analysis for llama-3.1-70b bdc91f5

complete gpt-4o-mini training 3f6b774

add env var END_REPETITION_PENALTY ee71b10

fine-tuned metrics WIP 2a89293

train with 4gpu a2100ac

ready for LF lora training 6fc6dc9

translation time analyzed 95ba169

complete results for gpt-4o and gpt-4o-mini fddc7fb

Update 00b_Data Analysis_Few_Shots.ipynb 60b04fa

refined data analysis 9a2b128

complete results for original prompt ad9aba8

ready for few shots prompting 4gpu 0156aec

eval with few-shots prompting 6c91c84

enable env: START_REPETITION_PENALTY a457091

completed Qwen2-72B results 7dab5c1

new mistral results analyzed 1bdff66

set max_new_tokens to 300 a37d279

formatted code 640a737

data analysis notebooks 1a9edc9

fixed bug for Mistral 12a5ff3

enable do_sample a69b127

set max_new_tokens to 2048 990ef6d

ready for gpu cluster c73d190

initial code for Chinese/English translation 3860729

v2e results

83aa28d

Update 03c_RAPGeT_v2_Data Analysis.ipynb

9ad0cbb

Chinese-English Machine Translation - Repetition Metrics for Different Models

a5c9876

analyzed llama-3-8b rpp1.0

0f06270

Update 03b_RAPGeT_v2_Data Analysis_Generic_Prompt.ipynb

d523de0

completed llama-3.1-70b generic prompt

db47db5

analyzed llama-3.1-70b rpp=1

6d425fc

added MTR graphs

f62fd81

ready for llama-3.1-8b

b83b354

chat template COMET

1ebd37e

COMET eval results

927a69c

few shot COMET results

101f384

eval comet with rtx4080

243d523

diable lfs for 00b notebook

553af0d

Update 00c_Data Analysis_Fine_Tuned.ipynb

80027ea

fine-tuned gpt-4o

46d9e0e

completed results: max_new_tokens = 2048 and new addition of Phi-3.5-mini (3.8B)

a35e6bd

ready to eval phi-3.5

25ab915

refine graph

b04757c

completed RPP analysis

0cc187e

completed analysis for llama-3.1-70b

bdc91f5

complete gpt-4o-mini training

3f6b774

add env var END_REPETITION_PENALTY

ee71b10

fine-tuned metrics WIP

2a89293

train with 4gpu

a2100ac

ready for LF lora training

6fc6dc9

translation time analyzed

95ba169

complete results for gpt-4o and gpt-4o-mini

fddc7fb

Update 00b_Data Analysis_Few_Shots.ipynb

60b04fa

refined data analysis

9a2b128

complete results for original prompt

ad9aba8

ready for few shots prompting 4gpu

0156aec

eval with few-shots prompting

6c91c84

enable env: START_REPETITION_PENALTY

a457091

completed Qwen2-72B results

7dab5c1

new mistral results analyzed

1bdff66

set max_new_tokens to 300

a37d279

formatted code

640a737

data analysis notebooks

1a9edc9

fixed bug for Mistral

12a5ff3

enable do_sample

a69b127

set max_new_tokens to 2048

990ef6d

ready for gpu cluster

c73d190

initial code for Chinese/English translation

3860729