init

experiments/plot.py

@@ -30,15 +30,15 @@ model_size = {
 lm_list = ['T5', 'Flan-T5', 'OPT', 'OPT-IML']
 
 # Oracle
-df_oracle = pd.read_csv("experiments/results/oracle.csv", index_col=0)
+df_oracle = pd.read_csv("results/oracle.csv", index_col=0)
 
 # Plot
-os.makedirs('experiments/figures/main', exist_ok=True)
+os.makedirs('figures/main', exist_ok=True)
 plt.rcParams.update({'font.size': 18})  # must set in top
 
 
 def main(target_relation: str = "average", prompt_type: str = "lc"):
-    df = pd.read_csv(f"experiments/results/lm_{prompt_type}/lm.csv")
+    df = pd.read_csv(f"results/lm_{prompt_type}/lm.csv")
     df.index = df.pop("model")
     df = (df * 100).round(1)
     df = df[[i in model_size for i in df.index]]
@@ -71,20 +71,20 @@ def main(target_relation: str = "average", prompt_type: str = "lc"):
                            grid=True,
                            figsize=(4, 5))
 
-    if prompt_type == 'qa':
+    if prompt_type == 'lc':
         ax.legend().remove()
     else:
-        ax.legend(loc='lower right')
+        ax.legend(loc='best')
     # ax.legend(loc='center left', bbox_to_anchor=(1, 0.5))
     plt.tight_layout()
-    plt.savefig(f"experiments/figures/main/{prompt_type}.{target_relation.replace(' ', '_').replace('/', '-')}.png", bbox_inches="tight", dpi=600)
+    plt.savefig(f"figures/main/{prompt_type}.is_{target_relation.replace(' ', '_').replace('/', '-')}.png", bbox_inches="tight", dpi=600)
 
 
 if __name__ == '__main__':
     for p in ['lc', 'qa']:
         main('average', p)
-        main("is competitor/rival of", p)
-        main("is friend/ally of", p)
-        main("is influenced by", p)
-        main("is known for", p)
-        main("is similar to", p)
+        main("competitor/rival of", p)
+        main("friend/ally of", p)
+        main("influenced by", p)
+        main("known for", p)
+        main("similar to", p)

experiments/plot_fewshot.py

@@ -4,10 +4,9 @@ import matplotlib.pyplot as plt
 import pandas as pd
 from random import shuffle, seed
 
-maps = {"is competitor/rival of": "Rival", "is friend/ally of": "Ally", "is influenced by": "Inf", "is known for": "Know", "is similar to": "Sim",
-        "average": "Avg"}
+maps = {"competitor/rival of": "Rival", "friend/ally of": "Ally", "influenced by": "Inf", "known for": "Know", "similar to": "Sim", "average": "Avg"}
 # Few-shots + Zero-shot
-os.makedirs('experiments/figures/fewshots', exist_ok=True)
+os.makedirs('figures/fewshots', exist_ok=True)
 plt.rcParams.update({'font.size': 16})  # must set in top
 # styles = ['o-', 'o--', 'o:', 's-', 's--', 's:', '^-', '^--', '^:', "X-", "X--", "X:"]
 styles = ['o', "X", '^', 'P']
@@ -16,16 +15,15 @@ colors = list(mpl.colormaps['tab20b'].colors)
 shuffle(colors)
 for prompt in ['qa', 'lc']:
     df = pd.concat([
-        pd.read_csv(f"experiments/results/lm_{prompt}_zeroshot.csv", index_col=0),
-        pd.read_csv(f"experiments/results/lm_{prompt}_fewshots.csv", index_col=0)])
+        pd.read_csv(f"results/lm_{prompt}_zeroshot.csv", index_col=0),
+        pd.read_csv(f"results/lm_{prompt}_fewshots.csv", index_col=0)])
 
-    df_full = pd.read_csv(f"experiments/results/lm_{prompt}/lm.csv", index_col=0)
+    df_full = pd.read_csv(f"results/lm_{prompt}/lm.csv", index_col=0)
     for r in maps:
         tmp = df[[r, "shot", "seed"]]
         tmp[r] = tmp[r] * 100
         ax = None
         for n, m in enumerate(['Flan-T5\textsubscript{XXL}', 'Flan-UL2', 'OPT\textsubscript{13B}', 'GPT-3\textsubscript{davinci}']):
-        # for n, m in enumerate(['Flan-T5\textsubscript{XXL}', 'OPT\textsubscript{13B}', 'GPT-3\textsubscript{davinci}']):
             g = tmp.loc[m]
             full_shot = df_full[[r]].loc[[m]] * 100
             full_shot["shot"] = 5
@@ -60,5 +58,5 @@ for prompt in ['qa', 'lc']:
         else:
             ax.legend(loc='lower right')
         plt.tight_layout()
-        plt.savefig(f"experiments/figures/fewshots/{prompt}.{r.replace(' ', '_').replace('/', '-')}.fewshot.png", bbox_inches="tight", dpi=600)
+        plt.savefig(f"figures/fewshots/{prompt}.is_{r.replace(' ', '_').replace('/', '-')}.fewshot.png", bbox_inches="tight", dpi=600)
 

experiments/plot_fewshot_landscape.py

@@ -4,10 +4,9 @@ import matplotlib.pyplot as plt
 import pandas as pd
 from random import shuffle, seed
 
-maps = {"is competitor/rival of": "Rival", "is friend/ally of": "Ally", "is influenced by": "Inf", "is known for": "Know", "is similar to": "Sim",
-        "average": "Avg"}
+maps = {"competitor/rival of": "Rival", "friend/ally of": "Ally", "influenced by": "Inf", "known for": "Know", "similar to": "Sim", "average": "Avg"}
 # Few-shots + Zero-shot
-os.makedirs('experiments/figures/fewshots', exist_ok=True)
+os.makedirs('figures/fewshots', exist_ok=True)
 plt.rcParams.update({'font.size': 16})  # must set in top
 # styles = ['o-', 'o--', 'o:', 's-', 's--', 's:', '^-', '^--', '^:', "X-", "X--", "X:"]
 styles = ['o', "X", '^', 'P']
@@ -16,16 +15,15 @@ colors = list(mpl.colormaps['tab20b'].colors)
 shuffle(colors)
 for prompt in ['qa', 'lc']:
     df = pd.concat([
-        pd.read_csv(f"experiments/results/lm_{prompt}_zeroshot.csv", index_col=0),
-        pd.read_csv(f"experiments/results/lm_{prompt}_fewshots.csv", index_col=0)])
+        pd.read_csv(f"results/lm_{prompt}_zeroshot.csv", index_col=0),
+        pd.read_csv(f"results/lm_{prompt}_fewshots.csv", index_col=0)])
 
-    df_full = pd.read_csv(f"experiments/results/lm_{prompt}/lm.csv", index_col=0)
+    df_full = pd.read_csv(f"results/lm_{prompt}/lm.csv", index_col=0)
     for r in maps:
         tmp = df[[r, "shot", "seed"]]
         tmp[r] = tmp[r] * 100
         ax = None
         for n, m in enumerate(['Flan-T5\textsubscript{XXL}', 'Flan-UL2', 'OPT\textsubscript{13B}', 'GPT-3\textsubscript{davinci}']):
-        # for n, m in enumerate(['Flan-T5\textsubscript{XXL}', 'OPT\textsubscript{13B}', 'GPT-3\textsubscript{davinci}']):
             g = tmp.loc[m]
             full_shot = df_full[[r]].loc[[m]] * 100
             full_shot["shot"] = 5
@@ -57,5 +55,5 @@ for prompt in ['qa', 'lc']:
         ax.set_xticks([0, 1, 3, 5])
         ax.legend(loc='lower right')
         plt.tight_layout()
-        plt.savefig(f"experiments/figures/fewshots/{prompt}.{r.replace(' ', '_').replace('/', '-')}.fewshot.landscape.png", bbox_inches="tight", dpi=600)
+        plt.savefig(f"figures/fewshots/{prompt}.is_{r.replace(' ', '_').replace('/', '-')}.fewshot.landscape.png", bbox_inches="tight", dpi=600)
 

experiments/results/lm_lc_fewshots.csv

@@ -1,7 +1,25 @@
 model,shot,seed,competitor/rival of,friend/ally of,influenced by,known for,similar to,average
+Flan-UL2,1,0,0.4634621318268161,0.16655626138700946,0.3560737650265537,0.41547148831050135,0.17042957362776867,0.3143986440357298
+Flan-UL2,1,1,0.4808325723430157,0.2700624443104191,0.3892318540513041,0.4232416819137535,0.2359978047515976,0.359873271474018
+Flan-UL2,1,2,0.3566701999849623,0.2570402256071756,0.35238039502018914,0.3265963193386597,0.21653807442945824,0.301845042876089
+Flan-UL2,3,0,0.6008806683724164,0.3409165910069645,0.4400867179850928,0.48946492285056153,0.3203158297994847,0.43833294600290396
+Flan-UL2,3,1,0.5984479531158479,0.3218657102958768,0.4100388352814914,0.5981330849874441,0.35945200787054604,0.45758751831024125
+Flan-UL2,3,2,0.5558370248323746,0.32140256084585034,0.479510078397158,0.5586568741198161,0.3275721270893454,0.44859573305690886
 Flan-T5	extsubscript{XXL},1,0,0.4289090393405381,0.2235622395277642,0.3116935132901571,0.24477064419482994,0.15137338610032808,0.27206176449072345
 Flan-T5	extsubscript{XXL},1,1,0.5408992995727435,0.28015138316349514,0.37012950513174664,0.22275827754974462,0.1943880588093273,0.32166530484541145
 Flan-T5	extsubscript{XXL},1,2,0.35487767295380657,0.19694658446624458,0.3643427270246087,0.15625343868751154,0.19403682711764508,0.25329145004996334
 Flan-T5	extsubscript{XXL},3,0,0.6051229823461516,0.3515458708850714,0.44012410027260657,0.3804674343263175,0.40166706800486446,0.43578549116700227
 Flan-T5	extsubscript{XXL},3,1,0.6207094316742009,0.3572426091203966,0.3608960801158353,0.49954040116534365,0.4290780859855123,0.45349332161225775
 Flan-T5	extsubscript{XXL},3,2,0.5559223832624296,0.3356752830641652,0.4745756164453349,0.399234265559725,0.41335486302169466,0.43575248227066987
+OPT	extsubscript{13B},1,0,0.4849127052996463,0.20449592050167562,0.2532949037360814,0.4419226019156754,0.25835706478379455,0.32859663924737464
+OPT	extsubscript{13B},1,1,0.48103743257514775,0.2688659748978508,0.29629201083446693,0.5712333238400675,0.2671453300905671,0.37691481444762004
+OPT	extsubscript{13B},1,2,0.41472246826539133,0.29268729494421086,0.3035815568996601,0.5357233481593575,0.2574976680913806,0.36084246727200003
+OPT	extsubscript{13B},3,0,0.6665554444567601,0.3765173453989971,0.4665832833748821,0.6225700574989006,0.5729112002367472,0.5410274661932575
+OPT	extsubscript{13B},3,1,0.6810749134091214,0.3487669741849127,0.47417188774018576,0.66063350589321,0.49442959968255856,0.5318153761819977
+OPT	extsubscript{13B},3,2,0.6545284416620056,0.445943447957961,0.5320695746415764,0.6374422462382947,0.5381467357749222,0.5616260892549519
+GPT-3	extsubscript{davinci},1,0,0.5280699275354719,0.212191920529615,0.3635277931568076,0.5348785543830825,0.3421445057868,0.3961625402783554
+GPT-3	extsubscript{davinci},1,1,0.4974091794597034,0.2292126628180868,0.39615505369886195,0.6084233413809402,0.29307669115731966,0.4048553857029824
+GPT-3	extsubscript{davinci},1,2,0.44272003332344273,0.1907326626783898,0.39534759628856364,0.613277325959876,0.27941602025529555,0.38429872770111356
+GPT-3	extsubscript{davinci},3,0,0.6604608525508305,0.35463353388524765,0.5785282815637409,0.6671818508708893,0.5031879729304646,0.5527984983602346
+GPT-3	extsubscript{davinci},3,1,0.6504483087053751,0.330626954058877,0.5698331614880284,0.7072260304355611,0.4307669873114737,0.537780288399863
+GPT-3	extsubscript{davinci},3,2,0.6291172370346219,0.3578601417204319,0.5625809977103491,0.671381955521296,0.4700152606032847,0.5381911185179967